Full Chapter Optical Inspection of Microsystems 2Nd Edition Wolfgang Osten PDF

Optical Inspection of Microsystems 2nd
Edition Wolfgang Osten

Visit to download the full and correct content document:
https://textbookfull.com/product/optical-inspection-of-microsystems-2nd-edition-wolfga
ng-osten/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Optical Inspection of Microsystems First Edition

Wolfgang Osten(Editor)
https://textbookfull.com/product/optical-inspection-of-
microsystems-first-edition-wolfgang-osteneditor/
Springer Handbook of Geographic Information 2nd Edition

Wolfgang Kresse
https://textbookfull.com/product/springer-handbook-of-geographic-
information-2nd-edition-wolfgang-kresse/
Elements of Optical Networking: Basics and Practice of

Glass Fiber Optical Data Communication 2nd Edition
Brückner
https://textbookfull.com/product/elements-of-optical-networking-
basics-and-practice-of-glass-fiber-optical-data-
communication-2nd-edition-bruckner/
Essentials of Modern Optical Fiber Communication 2nd

Edition Reinhold Noé (Auth.)
https://textbookfull.com/product/essentials-of-modern-optical-
fiber-communication-2nd-edition-reinhold-noe-auth/
Clinical psychology a modern health profession 2nd Ed
2nd Edition Wolfgang Linden
https://textbookfull.com/product/clinical-psychology-a-modern-
health-profession-2nd-ed-2nd-edition-wolfgang-linden/
Introduction to optical microscopy 2nd Edition Jerome

Mertz
https://textbookfull.com/product/introduction-to-optical-
microscopy-2nd-edition-jerome-mertz/
The Mexican Revolution s Wake Sarah Osten
https://textbookfull.com/product/the-mexican-revolution-s-wake-
sarah-osten/
Smart Sensors and MEMS Intelligent Devices and

Microsystems for Industrial Applications Woodhead
Publishing Series in Electronic and Optical Materials
1st Edition S Nihtianov A Luque
https://textbookfull.com/product/smart-sensors-and-mems-
intelligent-devices-and-microsystems-for-industrial-applications-
woodhead-publishing-series-in-electronic-and-optical-
materials-1st-edition-s-nihtianov-a-luque/
The Physics of Thin Film Optical Spectra: An

Introduction 2nd Edition Olaf Stenzel (Auth.)
https://textbookfull.com/product/the-physics-of-thin-film-
optical-spectra-an-introduction-2nd-edition-olaf-stenzel-auth/
Optical Inspection of
Microsystems
Optical Inspection of
Microsystems
Second Edition
Edited by
Wolfgang Osten
Cover image: Courtesy of Michał Józwik, Faculty of Mechatronics, Warsaw University of Technology.
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2020 by Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works
Printed on acid-free paper
International Standard Book Number-13: 978-1-4987-7947-0 (Hardback)
This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to
publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materi-
als or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material
reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained.
If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in
any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, micro-
filming, and recording, or in any information storage or retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.
com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a
not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a
photocopy license by the CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identifi-
cation and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at

http://www.crcpress.com
Contents
Preface to Second Edition.................................................................................................................vii
Preface to First Edition......................................................................................................................ix
Editor.................................................................................................................................................xi
Contributors.................................................................................................................................... xiii
Chapter 1 Image Processing and Computer Vision for MEMS Testing........................................1

Markus Hüttel
Chapter 2 Surface Features.......................................................................................................... 51

Xiangqian Jiang
Chapter 3 A Metrological Characteristics Approach to Uncertainty in Surface Metrology....... 73

Richard Leach, Han Haitjema, and Claudiu Giusca
Chapter 4 Image Correlation Techniques for Microsystems Inspection...................................... 93

Dietmar Vogel and Bernd Michel
Chapter 5 Light Scattering Techniques for the Inspection of Microcomponents and

Structures.................................................................................................................. 139
Angela Duparré and Sven Schröder
Chapter 6 Characterization and Measurement of Microcomponents with the Atomic

Force Microscope (AFM)......................................................................................... 155
F. Michael Serry and Joanna Schmit
Chapter 7 Optical Profiling Techniques for MEMS Measurement........................................... 177

Klaus Körner, Johann Krauter, Aiko Ruprecht, and Tobias Wiesendanger
Chapter 8 Grid and Moiré Methods for Micromeasurements...................................................207

Anand Asundi, Bing Zhao, and Huimin Xie
Chapter 9 Grating (Moiré) Interferometry for In-Plane Displacement and Strain

Measurement of Microcomponents........................................................................... 245
Leszek Salbut and Malgorzata Kujawinska
v
vi Contents
Chapter 10 Interference Microscopy Techniques for Microsystem Characterization................ 261

Alain Bosseboeuf, Philippe Coste, and Sylvain Petitgrand
Chapter 11 Measuring MEMS in Motion by Laser Doppler Vibrometry.................................. 297

Christian Rembe, Georg Siegmund, Heinrich Steger, and Michael Wörtge
Chapter 12 An Interferometric Platform for Static, Quasi-Static, and Dynamic Evaluation

of Out-of-Plane Deformations of MEMS and MOEMS.......................................... 349
Christophe Gorecki, Michał Józwik, and Patrick Delobelle
Chapter 13 Optoelectronic Holography for Testing Electronic Packaging and MEMS............. 379
Cosme Furlong
Chapter 14 Digital Holography and Its Application in MEMS/MOEMS Inspection................405

Wolfgang Osten and Pietro Ferraro
Chapter 15 Speckle Metrology for Microsystem Inspection...................................................... 485

Roland Höfling and Petra Aswendt
Chapter 16 Spectroscopic Techniques for MEMS Inspection.................................................... 517

Ingrid De Wolf
Chapter 17 Sensor Fusion in Multiscale Inspection Systems..................................................... 541

Marc Gronle
Index............................................................................................................................................... 563
Preface to Second Edition
2007 was a good year to publish a book dedicated to optical methods for the inspection of micro-
systems. The technological basis for the design and fabrication of microcomponents was meanwhile
mature, and many devices and applications were already implemented or under test. Consequently,
reliable procedures for quality assurance gained importance, and the search for efficient inspection
methods was on the agenda of many vendors. Full-field optical methods established an alternative
approach in comparison with conventional tactile and/or pointwise sampling methods, owing to their
noncontact, nondestructive, and areal working principle. Thus, this book was very well received by
the readers and translated to Chinese1 a couple of years after its first edition. However, 10 years later,
it was time to think about an update. As the technology progressed, naturally higher expectations
were placed on quality assurance and thus the inspection techniques used. For instance, challenges
such as checking the system’s internal parts after packaging came up. New requirements for the
calibration of measurement tools, new rules for the estimation of uncertainty in measurement, and
meaningful surface features had to be considered.
Consequently, the publisher and editor decided to prepare an updated and extended edition, tak-
ing into account the progress in the field. This second edition is now available. Three new chapters
were added, focusing on the following:
• Micro-Electro-Mechanical Systems (MEMS) surface features characterization

• Metrological characteristics to uncertainty in surface metrology
• Sensor fusion strategies for microsystems inspection
All these new chapters were written by leading experts in the particular fields. Nine of the original
chapters have been updated, while only five chapters have retained the content of the first edition.
Thus, the reader finds an almost completely updated and extended book. Publisher, authors, and
editor hope that the second edition will find an even well-interested audience and help to solve the
even more challenging problems in microsystems technology.
The editor is very grateful to all contributors and the publisher for the repeated fruitful coopera-
tion. Special thanks goes to Michelle Kazensky of Lumina Datamatics and Marc Gutierrez and
Kari Budyk at CRC Press.
Wolfgang Osten
Lilienthal, 2019
1 China Machine Press, Beijing, China, 2014, ISBN 978-7-111-45837-1.
vii
Preface to First Edition
The miniaturization of complex devices such as sensors and actuators is one of the biggest challenges
in modern technology. Different manufacturing technologies—for instance, the so-called LIGA
technique and UV lithography—allow the realization of nonsilicon and silicon microparts with a
high aspect ratio and structural dimensions in the range from nanometers to millimeters. LIGA is an
acronym standing for the main steps of the process, i.e., deep x-ray lithography, electroforming, and
plastic molding. These three steps make it possible to mass-produce high-quality microcomponents
and microstructured parts, in particular from plastics, but also from ceramics and metals at low
cost. Techniques based on UV lithography or advanced silicon etching processes (ASE) allow for
direct integration of electronics with respect to the realization of advanced microelectromechanical
systems (MEMS) devices. Further technologies such as laser micromachining, electrochemical
milling (ECF), electrodischarge machining, and nanoimprint lithography (NIL) offer, meanwhile,
an economical, high-resolution alternative to UV, VUV, and next-generation optical lithography.
Increased production output, high system performance, and product reliability and lifetime are
important conditions for the trust in a new technology and deciding factors for its commercial success.
Consequently, high quality standards are a must for all manufacturers. However, with increasing
miniaturization, the importance of measurement and testing is rapidly growing, and therefore the
need in microsystems technology for suitable measurement and testing procedures is evident. Both
reliability and lifetime are strongly dependent on material properties and thermomechanical design.
In comparison to conventional technologies, the situation in microsystems technology is extremely
complicated. Modern microsystems (MEMS and MOEMS) and their components are characterized
by high-volume integration of a variety of materials and materials combinations. This variety is
needed to realize very different and variable functions such as sensor and actuator performance,
signal processing, etc. Still, it is well known that the materials´ behavior in combination with new
structural design cannot be easily predicted by theoretical simulations. A possible reason for wrong
predictions made by FEM calculations with respect to the operational behavior of microdevices
is, for instance, the lack of reliable materials data and boundary conditions in the microscale.
Therefore, measurement and testing procedures are confronted with a complex set of demands. In
general, the potential for the following is challenged:
• Microscopic and nanoscopic measurement and testing on wafer scale

• Fast in-line measurement of various dimensional and functional properties of highly het-
erogeneous hybrid systems
• Verification of system specifications including geometrical, kinematical, and thermome-
chanical parameters
• Fast and reliable recognition of surface and subsurface defects, with the possibility for
review and repair
• Measurement of complex 3-D structures with high aspect ratio
• Determination of material properties well defined for the bulk but to be specified for
microscale
Measurement and inspection techniques are required that are very fast, robust, and relatively low
cost compared to the products being investigated. The reason for this demand is obvious: proper-
ties determined on much larger specimens cannot be scaled down from bulk material without any
experimental verification. Further on, in microscale, materials’ behavior is noticeably affected by
production technology. Therefore, simple and robust methods to analyze the shape and deformation
of the microcomponents are needed. Together with the knowledge of the applied load and appropri-
ate physical models, these data can be used for the derivation of material parameters and various
ix
x Preface to First Edition
system properties. It is obvious that neither a single method nor a class of measurement techniques
can fulfill these requirements completely. Conventional tensile test techniques (e.g., strain gauges)
are unable to test specimens from submillimeter-sized regions because of their limited local resolu-
tion and partly unwanted tactile character. Other approaches, such as, for instance, microhardness
measurements, do not reveal directional variations.
However, full-field optical methods provide a promising alternative to the conventional methods.
The main advantages of these methods are their noncontact, nondestructive, and fieldwise working
principle; fast response potential; high sensitivity and accuracy (typical displacement resolution of a
few nanometers, strain values of 100 microstrain); high resolution of data points (e.g., 1000 × 1000
points for submillimeter field of view); advanced performance of the system, i.e., automatic analysis
of the results; and data preprocessing in order to meet requirements of the underlying numerical
or analytical model. Thus, this book offers a timely review of the research into applying optical
measurement techniques for microsystems inspection. The authors give a general survey of the
most important and challenging optical methods such as light scattering, scanning probe micros-
copy, confocal microscopy, fringe projection, grid and Moiré techniques, interference microscopy,
laser Doppler vibrometry, holography, speckle metrology, and spectroscopy. Moreover, modern
approaches for data acquisition and processing (for instance, digital image processing and correla-
tion) are presented.
The editor hopes that this book will significantly push the application of optical principles for
the investigation of microsystems. Thanks are due to all authors for their contributions, which give
a comprehensive overview of the state of the art in the fascinating and challenging field of optical
microsystems metrology. Finally, the editor is grateful for the cooperation shown by CRC Press
represented by Taisuke Soda, Preethi Cholmondeley, Gerry Jaffe, and Jessica Vakili.
Wolfgang Osten
Stuttgart, 2007
Editor
Wolfgang Osten earned an MSc/Diploma in physics at the Friedrich Schiller University Jena in
1979. From 1979 to 1984, he was a member of the Institute of Mechanics in Berlin, working in the
field of experimental stress analysis and optical metrology. In 1983, he earned a PhD at the Martin
Luther University Halle-Wittenberg for his thesis in the field of holographic interferometry. From
1984 to 1991, he was employed at the Central Institute for Cybernetics and Information Processing
ZKI in Berlin, making investigations in digital image processing and computer vision. Between
1988 and 1991, he headed the Institute for Digital Image Processing at the ZKI. From 1991 until
2002, he joined the Bremen Institute of Applied Beam Technology (BIAS) to establish and direct
the Department Optical 3D Metrology. From September 2002 until October 2018, he was a full pro-
fessor at the University of Stuttgart and director of the Institute for Applied Optics. From 2006 until
2010, he was the vice rector for research and technology transfer of the Stuttgart University, and
from 2015 until 2018, he was the vice chair of the university council. His research work is focused
on new concepts for industrial inspection and metrology by combining modern principles of optical
metrology, sensor technology, and image processing. Special attention is directed to the develop-
ment of resolution-enhanced technologies for the investigation of micro- and nanostructures.
xi
Contributors
Anand Asundi Cosme Furlong
School of Mechanical and Aerospace Center for Holographic Studies and Laser
Engineering MicroMechatronics
Nanyang Technological University Mechanical Engineering Department
Singapore Worcester Polytechnic Institute
Worcester, Massachusetts
Petra Aswendt
ViALUX GmbH Claudiu Giusca
Chemnitz, Germany Cranfield University
Cranfield, United Kingdom
Alain Bosseboeuf
Centre de Nanosciences et de Nanotechnologies Christophe Gorecki
UMR CNRS 9001 Département LOPMD, FEMTO-ST
Université Paris-Sud Université de Franche-Comté
Université Paris-Saclay Besançon, France
Orsay Cedex, France
Marc Gronle
Philippe Coste Institut für Technische Optik
Centre de Nanosciences et de Nanotechnologies Universität Stuttgart
UMR CNRS 9001 Stuttgart, Germany
Université Paris-Sud
Université Paris-Saclay Han Haitjema
Orsay Cedex, France Mitutoyo Research Center Europe
Leuven, Belgium the Netherlands
Patrick Delobelle
Department LMA, FEMTO-ST Roland Höfling
Université de Franche-Comté ViALUX GmbH
Besançon, France Chemnitz, Germany
Ingrid De Wolf Markus Hüttel

imec, Department of 3D & Optical I/O Head of Department Machine Vision and
Technologies Signal Processing
and Fraunhofer Institute for Manufacturing
KULeuven, Department of Materials Engineering and Automation IPA
Engineering Machine Vision and Signal Processing
Leuven, Belgium Stuttgart, Germany
Angela Duparré Xiangqian Jiang

Fraunhofer Institute for Applied Optics and EPSRC Future Metrology Hub
Precision Engineering (IOF) Centre of Precision Technologies
Jena, Germany University of Huddersfield
Huddersfield, United Kingdom
Pietro Ferraro
Director of ISASI Michał Józwik
CNR - Institute of Applied Sciences Faculty of Mechatronics
& Intelligent Systems Warsaw University of Technology
Pozzuoli NA, Italy Warsaw, Poland
xiii
xiv Contributors
Klaus Körner Joanna Schmit

Institut für Technische Optik Principal Optical Engineer
Universität Stuttgart 4D Technology Business of Nanometrics
Stuttgart, Germany Tucson, Arizona
Johann Krauter Sven Schröder

Institut für Technische Optik Fraunhofer Institute for Applied
Universität Stuttgart Optics and Precision
Stuttgart, Germany Engineering (IOF)
Jena, Germany
Malgorzata Kujawinska
Institute of Micromechanics and Photonics F. Michael Serry
Warsaw University of Technology Veeco Instruments, Inc.
Warsaw, Poland Santa Barbara, California
Richard Leach Georg Siegmund

Professor of Metrology Polytec GmbH
Manufacturing Metrology Team Waldbronn, Germany
Faculty of Engineering
University of Nottingham Heinrich Steger
Nottingham, United Kingdom Polytec GmbH
Waldbronn, Germany
Bernd Michel
Micro Materials Center Berlin
Dietmar Vogel
Fraunhofer Institute for Reliability and
Micro Materials Center Berlin
Microintegration (IZM)
Fraunhofer Institute for Reliability and
Berlin, Germany
Microintegration (IZM)
Berlin, Germany
Wolfgang Osten
Institute of Electrical Information Technology
Tobias Wiesendanger
Clausthal University of Technology
Institut für Technische Optik
Stuttgart, Germany
Universität Stuttgart
Sylvain Petitgrand Stuttgart, Germany
Fogale Nanotech
Nîmes, France Michael Wörtge
Polytec GmbH
Christian Rembe Waldbronn, Germany
Institute of Electrical Information Technology
Clausthal University of Technology Huimin Xie
Clausthal-Zellerfeld, Germany School of Mechanical and Production
Engineering
Aiko Ruprecht Nanyang Technological University
Institut für Technische Optik Singapore
Universität Stuttgart
Stuttgart, Germany Bing Zhao
School of Mechanical and Production
Leszek Salbut Engineering
Institute of Micromechanics and Photonics Nanyang Technological University
Warsaw University of Technology Singapore
Warsaw, Poland
1 Image Processing and
Computer Vision for
MEMS Testing
Markus Hüttel
CONTENTS
1.1 Introduction............................................................................................................................... 2
1.2 Classification of Tasks...............................................................................................................2
1.3 Image Processing and Computer Vision Components..............................................................4
1.3.1 Behavior of Light, Colors, and Filters........................................................................... 5
1.3.2 Illumination................................................................................................................... 8
1.3.3 Lens Systems............................................................................................................... 12
1.3.4 Sensors......................................................................................................................... 14
1.3.4.1 CCD Sensor.................................................................................................. 15
1.3.4.2 CMOS Sensors.............................................................................................. 17
1.3.4.3 Color Sensors and Color Cameras................................................................ 18
1.3.4.4 Camera Types and Interfaces........................................................................ 19
1.3.4.5 Frame Grabbers............................................................................................20
1.4 Processing and Analysis of Image Data.................................................................................. 21
1.4.1 Computer Vision Process............................................................................................ 21
1.4.1.1 Image Acquisition......................................................................................... 21
1.4.1.2 Image Processing.......................................................................................... 22
1.4.1.3 Image Analysis............................................................................................. 22
1.4.1.4 Measurements and Decision-Making........................................................... 22
1.4.1.5 Presentation of Results.................................................................................. 22
1.4.2 Image Data Preprocessing and Processing Methods................................................... 22
1.4.2.1 Histogram..................................................................................................... 23
1.4.2.2 Point Transformations................................................................................... 23
1.4.2.3 Spatial Filtering............................................................................................26
1.4.3 Image Data Analysis Methods..................................................................................... 30
1.4.3.1 Point Operations........................................................................................... 30
1.4.4 Solving Measurement and Testing Tasks....................................................................34
1.4.4.1 Finding a Test Object or Region of Interest..................................................34
1.4.4.2 Position Recognition..................................................................................... 37
1.4.4.3 Measuring Geometric Features.................................................................... 39
1.4.4.4 Presence Verification.................................................................................... 42
1.4.4.5 Defect and Fault Detection........................................................................... 43
1.5 Commercial and Noncommercial Image Processing and Computer Vision Software........... 45
1.6 Image-Processing Techniques for the Processing of Fringe Patterns
in Optical Metrology........................................................................................................ 47
1.7 Conclusion............................................................................................................................... 48
References......................................................................................................................................... 49
1
2 Optical Inspection of Microsystems
1.1 INTRODUCTION
Not only is there a requirement for testing the electrical and dynamic behavior of MEMS, but there
is also a considerable demand for methods for testing these systems during both the development
phase and the entire manufacturing phase. With the aid of these test methods, it is possible to assess
such static properties as dimension, shape, presence, orientation, and the surface characteristics of
microsystems or their components. Using an optical measurement and testing technique based on
image processing and computer vision, a wide range of procedures can be applied, which enable
such properties to be recorded rapidly and in a robust and noncontact way.
If measurement and testing means are not based on special optical procedures but rather on
illumination with normal light and imaging with normal and microscopic optics, their resolution
capabilities extend only to just below the micrometer range. This is due to the diffraction of light
and the dimensions of imaging sensor elements in a lateral direction. Such a degree of resolution
is inadequate, as it is unable to cover the entire scale of microsystem structure sizes, which ranges
from just a few nanometers (e.g., surface roughness) to a few millimeters (e.g., external contours). In
order to measure sizes in the nanometer range, special imaging measurement and testing means are
required. These include interferometers, spectrometers, near-field/scanning electron/atomic force
microscopes (AFMs), and specialized reconstruction and analysis processes such as fringe process-
ing and scanning techniques, which are described in the following chapters.
The main advantages of implementing optical testing equipment using simple light sources and
normal and microscopic optics are the speed with which images can be recorded and analyzed and
the fact that they can be easily integrated into the manufacturing process, thus making the error-
prone removal of components from and reintroduction into the clean environment for test purposes
superfluous. For this reason, despite their limited resolution capabilities, they are still ideally suit-
able for testing large piece numbers, that is, in the manufacturing process in the areas of assembly,
function, and integrity testing and packaging.
Furthermore, the algorithms developed for image processing and computer vision are not only
suitable for analyzing images recorded using a video camera but can also be applied to the fields of
signal analysis, data analysis and reconstruction, and many other areas.
This chapter deals not only with the technical aspects of illumination and image recording
techniques but also with image processing and computer vision processes relevant to the optical
measurement and testing technique and their implementation in typical measurement and testing
tasks. Finally, several software products that are available commercially for image processing and
computer vision will also be described. However, first, a classification of typical measurement and
testing tasks in the field of microsystem development and production is given.
1.2 CLASSIFICATION OF TASKS

The tendency that has been observed for many years toward miniaturization in the electronics indus-
try has also been affecting the field of optics and mechanics over the last few years. The origins of
miniaturizing mechanical systems, with dimensions in the micrometer range, can be found in the
laboratories of research institutes concerned with semiconductor manufacturing and also in the labo-
ratories of the semiconductor manufacturers. They possess not only expertise in processing delicate
silicone structures but also the equipment required to produce them. Most of the microelectromechan-
ical systems or microoptoelectromechanical systems available today (so-called MEMS/MOEMS)
have been created using the combination of electronic, optical, and mechanical functional groups.
Today, MEMS/MOEMS can be found in a wide range of technical devices used in our everyday
lives. In ink-jet printers, MEMS-based microinjectors transfer the ink to paper. Digital micro mirror
devices (DMDs)—a matrix of thousands of electrically adjustable micromirrors—are responsible for
producing digital images in digital light projection (DLP) technology projectors. Sensors for measur-
ing pressure, force, temperature, flow rates, air mass, acceleration, tilting, and many other values are
Image Processing and Computer Vision for MEMS Testing 3
constructed as MEMS and are utilized especially in the automotive industry—a mass market with
high safety requirements (Airbag, Electronic Stability Program, Bosch, Germany [ESP®]). However,
MEMS and MOEMS are being used more and more in many other fields such as in the medical
industry or in biological and chemical diagnostics. The “Lab-on-a-Chip” is c apable of carrying out
complete investigation processes for chemical, biological, or medical analyses. Microspectrometers,
just a few millimeters in size, enable extremely small and cheap analysis devices to be constructed,
which can be used in flight equipment for monitoring terrestrial biological and climatic processes and
also in a wide range of technical applications in the form of hand-held devices. Using micromotors,
microdrives, micropumps, and microcameras, instruments can be constructed in the medical field for
keyhole diagnosis and surgical interventions.
Although, in comparison with microelectronic components, microsystems possess a clear three-
dimensional structure (e.g., microspectrometers, electrical micromotors, and gears for drives made
by using microinjection molding techniques), classic MEMS structures, especially sensor and mir-
ror systems, are essentially two-dimensional in shape. This feature is the result of constructing
MEMS based on semiconductor materials and on the manufacturing processes used in conjunction
with such materials. Another conspicuous characteristic of MEMS-based systems is the common
use of hybrid constructions, where the drive and analysis electronics are located on a semiconductor
chip and the actual MEMS (e.g., the sensor element) is located on a separate chip.
Both these properties influence the tests realizable for MEMS, using image processing and com-
puter vision. These are essentially performed using incident light arrangement, where both image
recording and illumination take place at the same angle. The transmissive light arrangement (which
can be much better controlled), where the object to be tested is situated between the illumina-
tion source and the image recording system, can be used to advantage if MEMS are more three-
dimensional in shape. This is becoming more and more the case.
From the point of view of image processing and computer vision, the solutions listed below are
possible for the following examples of testing tasks:
• Test object or regions of interest: Test objects need to be located if the position of the
objects to be tested varies either in relation to each another or in relation to the reference
system of the observation area. This is often the case with hybridly constructed systems,
where individual components are located relatively inaccurately next to one another.
Regions of interest need to be located, for example, in cases where MEMS components
are processed or adjusted individually (e.g., when measuring the cross-section of laser-
trimmed resistances and capacitors) or if a measuring area needs to be determined using
a small measurement range for a high-resolution measuring or test system (multiscaled
measuring/testing).
• Position recognition: The position of components needs to be recognized if they are
not aligned for a test or if their position, when installed, shows degrees of freedom (e.g.,
resistance, diode, and gear). Another example is when tests need to be carried out on the
components themselves, such as the recognition of component coding and the measure-
ment of geometric features. In some cases, it is necessary to merge the best two or more
images taken from the same scene (e.g., with different sensors or from different positions)
in one image. This very important task in image processing is called “image registration.”
There are a lot of approaches and methods to realize this “best fit” of images in medical,
remote sensing, and industrial applications [1].
• Measuring geometric features: Investigating a component from a metrological point of
view shows whether production tolerances have been adhered to; this affects both the
mechanical and electrical behavior of the object. The location of geometric features using
contour or edge recognition and the calculations based on them to obtain straightness,
circularity, length, enclosed surfaces, angle, distance, diameter, etc, form the principles of
optical metrology.
• Presence verification: The monitoring of production and assembly processes often requires
a purely qualitative statement regarding certain features, without specific knowledge of
their geometrical characteristics. For example, in production processes, tool breakage can
be monitored by checking a work piece for the presence of bore holes, grooves, etc. With
assembly processes, the focus of interest is usually on the assembled result, that is, the
presence of all components requiring assembly.
• Fault detection: In contrast with presence verification, in the case of fault detection, fea-
tures are checked for deviations from the required standards. To detect faults, for example,
text recognition is used for reading identification markings on components, color checking
for verifying color-coded components, and texture analysis for investigating structured
surfaces for flaws (e.g., scratches).
In order to solve these examples of measuring and verification tasks based on two-dimensional
image data, image processing and computer vision have a whole range of proven processing and
interpretation methods available. Shading correction, the averaging of image series, spatial and
morphological filtering, edge detection, pattern and geometric feature matching, segmentation, con-
nectivity analysis, and the metrology of geometric features denote some of these methods. Other
techniques that are required, for example, in interferometry, spectroscopy, and holography, and that
also permit imaging metrology, are described in the following chapters.
1.3 IMAGE PROCESSING AND COMPUTER VISION COMPONENTS

The elementary task of digital image processing and computer vision is to record images, to process/
improve and analyze the image data by using appropriate algorithms, to supply results derived from
them, and to interpret these results. Figure 1.1 shows a typical scenario for an image-processing
system. Images of an object illuminated by a light source are recorded by an electronic camera.
After having been converted by a frame grabber into digital values, the analog, electrical image
signals from the camera reach the memory of a computer. The digital representation of the scene
obtained in this way forms the starting point for subsequent algorithmic processing steps, analyses,
and conclusions, otherwise known as image processing and computer vision.
In addition to requiring the knowledge of the theory of digital image processing (which rep-
resents the fundamentals for the algorithms used in image processing and computer vision), it
becomes immediately clear from the scenario that expertise in the fields of light physics, optics,
electrical engineering, electronics, and computer hardware and software is also necessary. The
first steps towards finding a successful image-processing solution are not in the selection and
FIGURE 1.1 Components of a computer vision scenario.

implementation of suitable algorithms but rather much earlier on, that is, in the depiction of the
images. Particular attention must be given to the illumination of a scene. A type of illumination
well-adapted to the task in question produces images that can be analyzed using simple algorith-
mic methods. Correspondingly badly illuminated scenes may produce images from which even
the most refined algorithms are unable to extract the relevant features. It is also equally important
to take the dynamic behavior of imaging sensors into consideration. If illumination is too bright,
sensor elements may reach saturation range and produce overilluminated images, which could lead
to false conclusions. As far as metrological tasks are concerned, it is essential to understand the
imaging properties of lenses. For these reasons, the following sections are concerned with the most
important aspects of illumination, imaging sensors, camera technology, and the imaging properties
of lenses.
1.3.1 Behavior of Light, Colors, and Filters

The range of the electromagnetic waves (see Figure 1.2) that are of interest as far as image pro-
cessing is concerned lies between ultraviolet (UV) and near infrared (IR)—a very small part of
the entire electromagnetic spectrum. The human eye is capable of recognizing only a part of the
spectrum, that is, of visible light.
The intensity I [W/m2] of a light source is defined by the quantity of photons radiated per unit of
area. If the distance is increased between a spot-light source illuminating diffusely and a surface
element, the intensity of the light decreases proportionally to the square of the distance. The areas
of constant intensity are present as concentric spheres around the light source (see Figure 1.3).
As described later on, with light collimators, the conformity with this law leads to an inhomoge-
neous density of light, which decreases from the center toward the periphery.
Light from the sun or a light bulb is seen by the human eye as white light. This is due to the
fact that these light sources radiate electromagnetic waves that cover the entire spectrum of visible
light. As depicted in Figure 1.4, this becomes clear when light propagates through a prism made
of glass or similar transparent material. Owing to the higher optical density of glass, expressed as
FIGURE 1.2 Electromagnetic spectrum.
FIGURE 1.3 Illustration of the inverse-square law.
FIGURE 1.4 Refraction of white light by a prism.

the refraction index n, the electromagnetic waves of the light are refracted at different degrees,
depending on their wavelength λ, and disintegrate into the colors of the rainbow. This is known as
dispersion.
If light refracted into the colors of the rainbow is merged again using a lens, white light results.
This observation leads to the phenomenon of mixing colors. If lights of different colors are mixed, the
colors are added together and form white. If, however, colored substances (e.g., pigments) illuminated
by white light are mixed, the colors are subtracted from one another and form black (Figure 1.5).
From this, the following questions arise: Why objects appear colored if they are illuminated by
white light? What are the effects of colored lights in conjunction with colored objects?
If an object appearing red to the human eye is illuminated by white light, the electromagnetic
waves of the spectrum corresponding to the color red are reflected and reach the eye. The light
from all other wavelengths is adsorbed by the object’s pigments and is transformed into warmth.
Naturally, the same also applies for objects made up of several colors, as shown in Figure 1.6.
As monochrome cameras are often applied in image processing, colored light can be used advan-
tageously to highlight colored objects. If, as shown in Figure 1.7, a red-and-green-colored object is
illuminated by green light, essentially only the green-colored areas of the object reflect the light,
leading to pale gray values in an image taken by a monochrome camera.
FIGURE 1.5 Additive and subtractive colors.
FIGURE 1.6 Response of colored objects to white light.
FIGURE 1.7 Response of colored objects to colored light.

Today, light-emitting diodes (LEDs) are often used as colored light source in image
p rocessing if the area to be illuminated is not too large. Because of their specific color types,
LEDs cover the entire frequency range of visible light and the adjacent areas of near-IR and
UV, induce minimal loss of warmth, and can be switched on and off very quickly. However,
if intensive light sources are required, sources of white light such as halogen lamps are imple-
mented and combined with optical filters. The filters are generally made of colored glasses,
which selectively allow light of a certain wavelength to pass through and either adsorb or reflect
all other light wavelengths.
If a scene is illuminated with colored light but recorded with a monochrome camera, it often
makes sense in practice to place the filter directly in front of the camera lens rather than to filter
the light from the source of white light. In this way, any stray light from the environment that does
not possess the wavelength of the transmitted filter light is also filtered out and is therefore unable
to reach the camera’s sensors (see Figure 1.8).
In cases where the colors of an object are irrelevant, near-IR light sources are often used in
conjunction with a filter that transmits only this light. As a result, light conditions where the object
is illuminated are almost completely independent of visible light. If the IR light source is also
monochromatic, that is, only light from a narrow range of the spectrum is transmitted—or if the
filter is constructed as a narrow band pass (see figure 1.9)—the color distortions of the lens cause
fewer chromatic errors in the images recorded, which can be used advantageously, especially where
metrological tasks are concerned.
As far as the transmission of light is concerned, the behavior of optical filters is charac-
terized using spectral response plots. To do this, in general, the relative transmission of the
filter material is plotted as a function of the wavelength of the electromagnetic spectrum
(see Figure 1.9).
In the same way, it is also possible to characterize the spectral fractions of irradiated light from
a light source and the spectral sensitivity of light-sensitive sensors.
FIGURE 1.8 Effect of filter location on the resulting grayscale image.

FIGURE 1.9 Spectral response plot of: (a) edge filters, (b) band pass filters, (c) light sources, and (d) CCD
sensor.
1.3.2 Illumination
The type of illumination plays a crucial role in image processing. By using an illumination adapted
to the special task in question, the features of a test object requiring measurement or testing can
often be highlighted better. Thus, the processing of images is simplified drastically in many cases.
Illumination can essentially be classified into two types in accordance with the arrangement of
the light source, test object, and image-recording system. If the test object is situated between the
light source and the image-recording system, this is known as a transmissive light arrangement.
Correspondingly, if the light source and the image-recording system are situated on the same side
as the test object, this is known as an incident light arrangement.
The simplest realization of a transmissive light arrangement is represented by a light panel
(see Figure 1.10), in which the light source (usually consisting of several lamps, tube lamps, or LEDs)
(a) (b)
FIGURE 1.10 Transmissive light arrangement: (a) principle of a light panel and (b) silhouette of an object.
is positioned behind a ground-glass screen, which scatters light diffusely. This arrangement is
especially used to determine and measure the contours of flat, nontransparent test objects, because
the camera sees the shadowed image of the object. In cases where objects are translucent or trans-
parent, this arrangement enables internal structures to be recognized. However, the diffusely scat-
tered light from a ground-glass screen is disadvantageous if the dimensions of test objects are
particularly large in the direction of the optical axis of the camera. This is because surfaces that
are almost parallel to the axis could also be illuminated, thus falsifying the true silhouette of the test
object and the resulting measurement data obtained.
This disadvantage of diffusely scattered light can be reduced if an illumination system composed
of a concave lens and a spotlight source (see Figure 1.11)—a so-called collimator—is used instead
of ground-glass illumination. By placing the spotlight source in the focal point of the lens, the
light emerges as almost-parallel light rays. The LEDs (without lens optics) are especially useful as
light sources for this, because the light-emitting semiconductors are so small that they are almost
punctiform.
Despite the advantage of parallel light rays, illuminating collimators have the disadvantage of
producing inhomogeneous illumination. In accordance with the law regarding the decrease in light
intensity of a spotlight source, the light intensity of light rays passing through the lens decreases
from the optical axis toward the periphery, known as shading or vignetting. This effect, which is
disadvantageous as far as image processing is concerned, can be avoided if a homogeneously illu-
minated diffuse ground-glass light source, placed at the image level of a lens, is projected, instead
of using a spotlight source in the focal point of a collimator. If a telecentric lens (see Section 1.3.3)
is used instead of a normal lens to project the ground-glass light source, homogeneous illumination
with parallel light rays results.
As many MEMS or their components are based on silicon wafers, which are—according to their
nature—essentially two-dimensional in shape, the main type of illumination used in image process-
ing to test such elements is that of incident light arrangement. In contrast with transmissive light
arrangement, by using this type of illumination, both the test features of the object and all other
areas are equally illuminated. This results in the fact that features in the images being recorded are
often more difficult to differentiate in the subsequent image processing. From this point of view, a
type of illumination that is adapted to the task is of particular importance.
With incident light arrangements, an essential difference is made between dark field arrange-
ment and bright field arrangement, because they highlight very different aspects of a test object.
As far as the angle of observation is concerned, with dark field arrangement (see Figure 1.12),
the test objects are illuminated from an almost-perpendicular angle. In an ideal situation, as a result,
no or only a small amount of light falls onto the object surfaces, which are parallel to the angle of
(a) (b)
FIGURE 1.11 Transmissive light arrangement: (a) collimator and (b) projection of a homogeneous light
source.
(a)
(b)
FIGURE 1.12 Principles of dark-field illumination: (a) single sided and (b) circular light.
incidence and which are thus perpendicular to the angle of observation. These fields are seen by
the observer as being dark areas. In contrast, all other nontransparent object details are highlighted,
provided they are not positioned in the shadows of other object details. With nontransparent test
objects, dark-field illumination is therefore advantageous if raised sections of an object have to be
highlighted. Another example of using dark-field illumination to advantage is for detecting particles
on smooth, even surfaces. Transparent test objects such as glass bodies and plastics can also be
examined for inclusions or edge defects, as these features stand out well against a dark background,
especially when the latter is matt black in color. Generally, spot and linear, diffuse and directed
light sources are used for dark-field illumination. In many cases, low-angle glass fiber or LED ring
lights reduce the problem of shadow formation.
In contrast with dark field arrangement, in the case of bright field arrangement test, objects are
illuminated almost from the angle of observation. In simple cases, the light source for bright field
arrangement is realized using either one or several spotlight sources, high-angle glass fiber lights,
or LED ring lights placed around the camera lens. Using this illumination arrangement, a good dif-
ferentiation of features can be obtained where object surfaces are diffusely dispersive, provided the
colors of the features to be tested contrast well against their surroundings, such as in the case of dark
print markings and barcodes on pale or colored markings.
The bright field arrangement described earlier becomes problematic if object surfaces are shiny
or reflective. This is because the light sources may be projected in the recording camera system in
the form of reflections, thus impeding or preventing reliable feature analysis. Homogeneous illu-
mination can be achieved in these cases if all visible sides of an object are illuminated equally by
a diffuse light source. The light dome (half sphere) shown in Figure 1.13 is capable of doing this.
Light sources fixed at the edge of the light dome illuminate its internal surface, which is matt and
coated white.
In the case of flat, shiny surfaces, reflex-free bright-field illumination can be achieved if the
light is coupled coaxially to the angle of observation. As shown in Figure 1.14a, diffuse light from
a homogeneously illuminated surface—such as that used in transmissive light arrangements—is
deflected by 90° toward the angle of observation by using a semitransparent mirror. Owing to the
(a)
(b)
FIGURE 1.13 Bright-field illumination: (a) spotlight source and (b) light dome.
(a)
(b)
FIGURE 1.14 Bright-field illumination: (a) coaxially diffuse and (b) coaxially directed.
homogeneity of the light source and the evenness of the object surface, object areas possessing
an equal degree of reflection are depicted in the camera image with the same level of brightness,
making this a reliable method for imaging and analyzing features on such surfaces.
A special form of coaxial bright-field illumination is utilized for telecentric lenses
(see Section 1.3.3). As shown in Figure 1.14b, light from a spotlight source is coupled to the beam
path of the lens by using a semitransparent mirror. Ideally, the light emitted from the lens is made up
purely of parallel light rays, which are reflected toward the telecentric lens from object surfaces only
if those surfaces are perpendicular to the optical axis. The properties of telecentric lenses prevent
the light of all other surfaces from reaching the camera, with the result that these appear as dark
areas in the recorded image. This feature can be used to make the edges of flat, three-dimensional
structures visible.
Besides the types of illumination described here, other forms of illumination can also be imple-
mented in image processing. These include types of structured illumination such as fringe projec-
tion and laser scanning for recording and measuring objects in three dimensions (see Chapter 6).
1.3.3 Lens Systems

To project a real scene on the imaging sensor of a camera, a wide range of various lenses is avail-
able. Depending on the task, normal, macro-, and microscopic lenses can be used. In principle, the
imaging properties of these lenses do not differ much from the imaging behavior of a single thin
lens shown in Figure 1.15, with the imaging equation:
1 1 1
= + (1.1)
f i o
As can be seen in Figure 1.15a, the geometric construction of the image Im results from the object O
as follows: a line is drawn from each object point through the center of the lens; a second line paral-
lel to the optical axis is bisected with the principal plane of the lens; and starting at this intersection,
a third line is drawn through the focal point of the lens. The intersection of the first and third line
then gives the corresponding image point.
As shown in Figure 1.15b, it can be seen from the lens equation that both the position and the size
of the image Im alter if the object is moved along the optical axis. In the case of lenses, in order to
depict the object in focus, the displacement of the image Im needs to be corrected. This is achieved
by moving the lens along the optical axis, so that the focal plane (where the imaging sensors of the
camera are situated) is always in the same place. Owing to the limitation of the depth of field, it can
be deduced directly that the objects that are particularly large in the direction of the optical axis and
that are close to the lens cannot be depicted completely in focus. To be more precise, this results in
part of the image to be in focus and the surrounding areas to be slightly out of focus. This charac-
teristic may considerably impair metrological image analysis or even render it impossible.
The cause of this blurred imaging can be explained by the following observation (see Figure 1.16a).
Each point of the object O emits light homogeneously in all free directions in the form of spherical
waves. Part of the light reaches the lens, and this then depicts it as an image point in the zone of
sharp focus. Circles of confusion are formed outside the zone of sharpness instead of image points,
thus resulting in a blurred image.
(a) (b)
FIGURE 1.15 Imaging behavior of a thin lens: (a) geometric construction of the image of an object, and
(b) if an object moves toward a lens, the image moves away from the lens and gets enlarged.
(a) (b)
FIGURE 1.16 (a) Formation of image points from object points and (b) telecentric imaging.
As the geometric construction shows, the light rays running parallel to the optical axis possess
two characteristic features. Independent of the distance between the object and the lens, they always
run parallel, and when they are projected, they always pass through the focal point F of the lens.
This results in the concept of generating a sharp image, independent of distance, by utilizing a
screen with a small aperture at the height of the focal point (focal pupil) in the beam bath of the lens
(see Figure 1.16b). Using this measure, a so-called object-sided telecentric lens is created.
As illustrated in Figure 1.17, telecentric lenses possess the following properties:
1. The field of view is equal to the size of the lens, because light rays running parallel to the
optical axis are not imaged outside the lens area.
2. The imaging of an object O (see Figure 1.17a) always results in (within limits) a sharp
image Im, independent of the distance between the object and the lens. This characteris-
tic is limited by the finite dimensions of the focal pupil, because almost no light passes
through an infinitely small point.
3. For all distances from the lens, the object is imaged with the same magnification as far as
the focal plane remains fixed.
4. Owing to the focal pupil, telecentric imaging is of low light intensity.
5. The displacement of the focal plane (see Figure 1.17b) along the optical axis alters the size
of an image but not its sharpness.
(a)
(b)
(c)
FIGURE 1.17 Behavior of one- and double-sided telecentric lenses: (a) within limits, object movement
always result in a sharp image; (b) within limits, movement of the focal plane changes size of image but
not sharpness; and (c) in case of double-sided telecentric lenses, sharpness and size of image (within limits) do
not alter if focal plane moves along the optical axis.
This characteristic predestines telecentric lenses to being used for tasks where precise metrological
image analysis is required.
The transition from an object- to a double-sided telecentric lens is achieved by introducing a
second lens behind the focal pupil (see Figure 1.17c) in such a way that the focal points of the two
lenses coincide. As a result, the focal rays are imaged into rays running parallel to the optical axis,
and subsequently, the alteration in image size, which takes place when the focal plane is moved
along the optical axis, no longer occurs. Double-sided telecentric lenses can be used to a dvantage
if imaging sensors are equipped with microlenses (one lens per pixel) to give a better light yield.
In this case, light entering obliquely may lead to vignetting, thus resulting in inhomogeneous
grayscale value distributions in the images recorded.
1.3.4 Sensors
Imaging sensors built into today’s cameras are based on semiconductor materials. The formation of
“electronic” images by using such sensors utilizes the principle of the internal photo effect. With
this principle, light quanta (photons) possessing a minimum material-specific energy release atomic
electrons from their bonded state (see Figure 1.18a). However, in the process, the freed electrons
remain in the material.
The external photo effect describes the process where electrons are emitted from a material.
Photomultipliers (not discussed here) are based on this effect, for example.
The light-sensitive elements of imaging sensors based on semiconductor materials are con-
structed as photodiodes and use the depletion-layer photo effect. The functioning principle of such
diodes is explained briefly below.
A silicone substrate is p-doped with boron atoms; owing to the three free valence electrons
of boron, there are insufficient electrons to bond with the four valance electrons of silicone. The
substrate is then coated with a thin layer of silicone n-doped with fluoride. As fluoride atoms pos-
sess five free valence electrons—that is, one electron more than necessary in order to bond with
silicone—some of the free surplus electrons fill in the missing electrons (holes) in the p-substrate
of the junction zone (see Figure 1.18b). Through this recombination, the p-substrate in the junction
zone becomes negatively charged and the n-doped silicon layer positively charged, thus preventing
additional electrons from migrating from the n-doped silicone to the p-doped silicone. The junction
zone is thus transformed into a depletion layer.
(a) (b)
FIGURE 1.18 Illustration of the principle of (a) the photo effect and (b) of the photodiode.
Another random document with
no related content on Scribd:
Fig. 18.—Modern.
Fig. 19.—Typical Grain-Handling Plant; 50 Tons per hour.
B is a length of flexible pipe to enable the nozzle to sweep the

width of the barge, the length being dealt with by moving the boat to
within the radius of the swivel arm E.
C is a permanent pipe carried on the jib arm which is mounted on
a swivel joint; B is another short flexible pipe to permit of the
swivelling already mentioned.
F indicates the receiver into which the grain is deposited, and G
the special rotary valves, for releasing the material from the system
without interfering with the vacuum of the conveying line.
H illustrates the twin air filters which are provided with valves, so
that each of the filters can be isolated in turn from the rest of the
plant for cleaning of the fabric tubes, without interrupting the
continuous working of the plant.
J indicates the connecting pipes between the receiver, air filters,
and pump, and it should be noted they are considerably larger than
those on the suction end C, this being necessary to allow for the
expansion of the air under vacuum.
K is the reciprocating air pump, driven electrically by the motor M
through large helical gear wheels, which act as flywheels and are
helpful in equalizing the torque required by the pump.
L is the outlet pipe for the air as it is exhausted from and by the
pump.
Fig. 20.—Floating Pneumatic Transport Plant; 200-250 Tons

per hour.
Fig. 21.—Portable Pneumatic Plant on Railway Truck.
This plant is simple, but every possible modification can be added

that circumstances make desirable; such as discharging on to a belt
conveyor, for feeding the silos or other storage bins.
Where grain is bought by bulk it would be necessary to check the
weight, and in this case an automatic weigher would be installed,
into which the material would be discharged from the receiver. With a
bag filter the collected dust can also be weighed, the total weight
being then obtained by addition.
Typical Floating Plant. The most popular development of the
above type of plant is the floating plant, designed for erection on a
hulk, or barge, and intended to suck grain from an ocean-going
steamer and discharge it into barges or lighters on the other side of
the discharging plant, as indicated in Fig. 20. The plant shown is one
of many in actual operation and its construction may be followed by
reference to the description of the plant shown in Fig. 19, the same
index letters being used in each case. One great advantage of
floating plants is that the difference in height due to tides is of no
consequence.
Portable Plants. Still another modification consists of a complete
quay-side plant mounted like a travelling crane, so that it can be
moved longitudinally along steel rails on the quay-side. This plant is
particularly useful where large cargo ships have to be emptied and
then allowed to remain at their berth until reloaded with another
cargo.
Fig. 22.—Portable Railway Plant in Operation.
Complete plants have also been mounted on railway trucks, the

engine and pump in this case being on a second waggon (see Figs.
21, 22). In such a case the plant has to be mounted very low, and it
is necessary to lift from the under-side of the truck to, say, a railway
waggon, by means of an adjustable belt or bucket elevator (shown at
the extreme left of Fig. 21).
Many other applications will suggest themselves to the reader, and
sufficient has been said to prove that for the handling of wheat the
pneumatic system is distinctly flexible and convenient; also, it effects
a great saving in labour, which is an important consideration
nowadays. “Bushelling,” conveying, and weighing by hand used to
cost well over a shilling per ton, which figure was reduced to just
over 1½d. per ton by pneumatic conveying; these are pre-war costs
in both instances, but the relative saving effected by pneumatic
conveying is certainly not less under present conditions.
All the previous remarks also apply to linseed and cotton seed in
bulk, maize, oats, and in fact all cereals. Such materials have to be
accepted as and when the ships arrive irrespective of convenience,
and it is an important advantage of pneumatic conveying that the
material can be lifted and discharged in the most convenient
position; also, when the barge or ship has departed the same
apparatus can be utilized to lift the material from its position in store
to the cleaning or grinding plant.
CHAPTER V
PNEUMATIC COAL-HANDLING PLANTS
The writer was directly interested in the erection and installing of one
of the first plants installed in this country for the elevation and
conveying of coal, and a description of the various details may give a
good idea of a complete plant, handling coal on a commercial scale.
The conditions to be complied with are as follows: 20 tons of
“slack” per hour, to be raised 90 ft. above canal level or 80 ft. above
road level.
The coal is brought alongside the power-house by canal barges of
25 tons capacity, or by tipping steam waggons from the railway
sidings, a distance of one mile away. In both cases the coal is
required to be elevated into overhead bunkers of 600 tons capacity
placed vertically over the boilers.
Fig. 23.—Pneumatic Unloading of Coal at Messrs. Boots, Ltd.
(Nottingham).
In the first case, immediately the barge is alongside, the flexible

suction pipe is lowered into the barge (Fig. 23), and coal immediately
begins to rise in the pipe and is discharged as required. Little or no
handling of the coal is required after the suction nozzle has once
reached the bottom of the barge; all that it is necessary to do is to
bring the barge gradually up to the nozzle, the coal then
“avalanching” down to the nozzle. Fig. 24 shows the discharger
placed on girders over the bunkers into which it discharges
continuously. The coal enters at A and the major portion of the fuel is
discharged through the rotary valves B.
The coal discharger itself consists of a cast iron vessel with two
King’s patent rotating valves. These are designed in the form of a
slightly conical taper divided into four sections, one portion of the
circular valve being under vacuum, and the other under ordinary
atmospheric pressure. The outlet of the valves is larger than the inlet
to allow the coal which is in the valve to drop out easily. Over each
valve is also provided a four-armed sweeper to prevent any damp
coal from forming a cone inside. The discharger is provided with two
inlets with full-way bored valves, so that the coal can be drawn either
from the water side or from the land side at will. After the coal has
been deposited in the main discharger, there is provided a
supplementary discharger consisting of a vessel 6 ft. high by 30 ins.
diameter, with two inlet pipes of 8 ins. diameter, to provide a contra-
flow, so that any particles of coal dust in the air will meet one another
in the 30 ins. box at equal velocity and be deposited. The small
particles are delivered by a supplementary rotary discharge valve
which is set to run very much slower than the main discharger
valves.
Fig. 24.—Discharger for Coal Conveying Plant.
The main discharger valves are driven by worm gearing, the latter
having one right-hand thread and one left-hand thread, so that the
end thrust on the worms is neutralized. Ball bearings are provided
and the small motor which drives all three valves is coupled up with
an electrical device designed by the author. This device ensures that
if anything happens to the top discharge valves—so that the 3 h.p.
motor driving them cuts out, owing to an overload or other cause—
then the main motor also is cut out by the opening of its circuit
breaker. This prevents any “flooding” of the pipes and dischargers.
It may be mentioned that the valves are so designed that a portion
of the weight of each valve is carried by the vacuum, so that the
vertical wearing lift on the valves when at work is very slight.
The intake pipes for the coal are 5 ins. diameter, and they are
provided with heavy cast iron bends, having extra thick metal on the
outside radius to allow for the wearing effect of coal passing at the
rate of 20 tons per hour.
The pipe into the barge is provided with a flexible steel pipe at the
suction nozzle end, for convenience of handling. India-rubber piping
has been tried, but the extra cost does not justify its continued use.
The nozzle is made as light as possible for convenience of
handling, and is fitted with a special “free air” inlet for the regulation
of the amount of air necessary to blend with the coal.
Ash Handling. In addition to unloading coal, the above plant is
capable of dealing with hot ashes which are first crushed in a
portable clinker breaker, electrically driven, which runs under all the
ash hoppers of the boilers. The ash when crushed gravitates into
funnel-topped tee-pieces, inserted in the main ash-conveying pipe,
whence it is immediately sucked up into an overhead ash hopper to
await the convenience of the waggons which dispose of it on the
“tips.”
Flue Cleaning. A 3 in. suction pipe has been run round the boiler-
house in such positions that flexible hose can be attached for flue
cleaning purposes. In this case the cleaners simply use an enlarged
nozzle such as is supplied with a domestic equipment and the dust is
removed from the flues, economizer soot chambers, etc., into the
ash hopper without trouble or dust.
The success of this plant is best indicated by the fact that, at the
moment of writing, a duplicate plant is being erected. Owing to the
growth of the business, and its demand for power and steam, the
original plant has to be worked continuously on coal, so that the ash
and flue dust problem has become acute again.
Portable Floating Plant. A third plant ordered by the same firm is
of considerable interest. This is intended to be mounted in a barge
so as to be portable. Owing to lack of space in close proximity to the
power-house, considerable difficulty is found in keeping adequate
stocks of coal on the site except the 600 tons in the overhead
bunkers. In order to secure continuity of working, it is essential that
as much fuel as possible be stored, and for this purpose a coal pile
has been made about half a mile away from the works, adjoining the
canal. Ashes can be disposed of on certain fields a few miles outside
the city in swamps and pools caused by subsidences, due to colliery
workings.
The portable plant is therefore arranged to operate as follows: the
barge is self-propelled by a 30 h.p. paraffin engine which can be
coupled by clutches to either the propeller or a Roots blower, the
latter being the exhauster for the portable suction plant.
The barge is loaded with ashes for disposal, and then proceeds
under its own power to the site where they are to be dumped. The
clutch is operated disconnecting the propeller and operating the
blower. The suction side of the blower is coupled up with the pipe
line in the boat and the barge feeds the plant by means of the
flexible hose: the discharge pipe is raised over the towing path so as
not to interfere with passing traffic, and the ashes are blown out into
the swamps previously mentioned. It will readily be recognized how
simple this unloading becomes compared with trying to dig out the
ashes with either a spade or a fork.
The empty barge then returns to the coal pile and takes up a load
of coal in a similar manner, then proceeding to the power-house
under its own power and being unloaded by the original fixed
pneumatic installation in the ordinary way.
The coal arriving by road is tipped into a concrete hopper
excavated below the ground level, and so designed with sloping
sides that it is self feeding into a suction pipe connected to the
bottom of the hopper. The same procedure occurs except that in this
case the coal enters the main discharger at the top (E, Fig. 24).
It is interesting to note that the very fine dust collected from the air
filter is eagerly sought after by the foundry trade, and what would at
first appear to be a waste product impossible to burn, is actually a
valuable by-product of the plant.
CHAPTER VI
THE INDUCTION CONVEYOR
Numerous means have been devised to cause the necessary current

of air to flow along the conveyor pipe, but the ideal method is
probably yet to seek. Probably the most satisfactory and economical
system, until recently, was the positive pump exhausting a vacuum
chamber; the latter receiving the material, and discharging it into the
receptacle provided for that purpose.
The difficulties arising in practice, however, incited the inventive
genius of engineers responsible for the operation of these plants,
and a number of attempts were made to induce an air current by
other means.
Ejector Systems. Steam ejectors were fitted to the closed tank
provided for the reception of the material, thus converting the tank
into a vacuum chamber, and eliminating the discharger. In other
cases injectors, also operated by steam, were placed at intervals
along the conveyor pipe, usually at such convenient points as 90°
bends, and the slight vacuum created by the condensation of the
steam and also by the velocity of the jet, induced an air current
which swept the material along with it into the receiver chamber.
Although both these methods are in practical use, their
applications are strictly limited to materials which do not suffer by
contact with heat and moisture; the methods are therefore used
principally for conveying ashes and soot from boiler furnaces and
flues. Ashes formed by the combustion of coal contain large
amounts of abrasive matter, and it is very important that all this
matter should be extracted from the air, before entering the
exhauster of the suction system. The steam jet cuts out the
exhauster entirely, but absorbs an excessive amount of power in the
form of steam. It has the advantage, however, of quenching the
ashes on their way to the settling tank. For flue dust, however, the
steam jet is unsuitable, as the condensed vapour causes the
material to cake in the pipes, and the latter rapidly become choked,
involving considerable delay and trouble in cleaning out. The ejector
system is used for this material, the tank being of the closed type,
and the necessary vacuum being created by a steam ejector fixed in
a branch at the side near the top, the dust striking a baffle and falling
by gravity to the bottom of the tank. A special air-tight gate or valve is
opened to empty the tank.
Air Induction. The cardinal feature of the induction system is the
ease with which materials may be handled which cannot be
conveyed by the suction method.
Sand, sugar, salt, soda ash, and many other substances of a
granular nature, which are very troublesome when conveyed by the
suction method, may be dealt with economically by the induction
system and, although the latter is only in its experimental stages at
the moment of writing, it is possible that it may displace all other
systems in the near future.
The induction system differs from the suction system in that the air
flow along the pipe is induced by a jet of air, at very high velocity,
fixed at any convenient distance from the intake nozzle of the
conveyor pipe, and the material conveyed is discharged either from
an open end into an open container, or by some form of cyclone. The
closed discharger or container, with its baffles and rotary valves or
air locks, is eliminated, and the substance to be handled has a free
and unrestricted flow throughout the length of the pipe.
Advantages of the Induction System. The advantages of the
induction conveyor may be summarized as follows: (1) Low first cost,
the power unit being the only expensive item. (2) Low maintenance
cost, there being no moving parts and little wear. (3) Low labour cost,
practically no attendance being required. (4) Flexibility and ease of
handling. (5) High efficiency of power unit and reliability of system.
(6) Ability to handle materials which are easily damaged.
In the case of a suction plant handling grain or coal, the intake end
of the conveyor is fairly flexible, and the nozzle may be operated
over a fairly large radius, say, all over the floor of a vessel’s hold.
The discharge end, however, is fixed, unless a cumbersome and
expensive gantry is provided to permit of the discharge apparatus
being moved about. Even should the discharger be mounted on rails,
the area over which it can operate is limited by the rails on which it
runs. With a large plant, this would mean that a number of
dischargers would be required to lift from a ship into a warehouse or
store, from the latter into bunkers or silos, or perhaps into trucks or
waggons. Each discharger would require a separate exhauster and a
separate intake, and valuable space would be occupied by the plant
and expense incurred for machinery which would not be in use for a
considerable part of the time.
The induction system, however, is flexible at both intake and
discharge ends of the pipe. It is only necessary to lower the nozzle
into the material to be removed, and to place the delivery pipe over
the receptacle for the material, and to turn on the air jet. The delivery
may be handled easily while working, and the material distributed
where required; or suitable valves and branches may be fixed, and a
number of discharge pipes used in turn to deliver into different bins
or into various floors.
The source of power for the operation of the induction conveyor is
the air compressor. As every operating engineer is well aware, all
machinery is kept in better condition and runs more economically,
when it is housed in proper environment and receives skilled
attention. The compressor, in this case, need not be erected near the
work, but may be placed some distance away, preferably in the
power station, as the pipe line connecting the conveyor with the
machine will have a very small bore, compared with the air pipe to
the exhauster on a suction plant, and will also be inexpensive to
erect and maintain. In the instance quoted above, where a number of
suction and delivery points are required, only these small pipes need
be run from a common main, and turned on and off as needed, the
compressor running continuously at or near its most economical
load.
Construction of Induction Plant. The induction conveyor may be
said to be a compromise between the suction and blowing methods.
The air jet is fixed in the conveyor pipe at a suitable angle, some
distance above the intake nozzle, and a stream of air at high velocity
is passed along the pipe in the direction of the discharge. This air jet
is designed carefully for the duty it has to perform, and its discharge
entrains the free air in the pipe, causing it to move in the required
direction. A partial vacuum is created in the conveyor pipe, behind
the jet, and free air rushes in at the intake, carrying the material
along with it.
In order to effect the greatest economy in the operation of this
plant, it is important that the power unit should be carefully chosen,
and that the pipe system should be designed to give the full pressure
at the jet. With modern multiple-stage compressors of the rotary or
reciprocating type, working at about full load, very high efficiencies
can be obtained, and the pressure pipe line should be arranged to
avoid loss by friction as far as possible. The receiver should be of
sufficient capacity to absorb any pulsations, and to throw down oil
and moisture before the air enters the pipes. A separator of good
design should also be incorporated.
In designing the conveyor pipe line, bends should be avoided
when possible, by erecting the lifting pipes at an angle with the
horizontal. It is not sufficiently well appreciated that bends and
angles rapidly increase the frictional resistance to the flow of the
conveying medium, and mean loss of power; in fact there is no doubt
that the difference between success and failure in pneumatic
conveying is largely a matter of design. Many substances which are
otherwise quite suitable for handling in this manner are very fragile,
and any friction on pipe walls or contact with metal baffles at high
velocity, so reduces or pulverizes them that their value is reduced
considerably. In the case of ashes from boiler furnaces, this effect is
advantageous rather than otherwise, but when dealing with coal it is
necessary to arrange the system so that the minimum amount of
damage is done to the material. Some coals, such as Derbyshire
bituminous, is not easily broken or abraded, and can be lifted very
satisfactorily by the usual suction method. Welsh coal, on the other
hand, is very friable, and if conveyed into the usual discharger, will
emerge in a finely divided state, even though it may be fed to the
intake in large pieces. For handling such materials, the induction
method is most suitable, as the discharge end may be arranged so
that the material is not delivered at high velocity, and does not strike
any obstacle which would reduce it or break it up. It is possible to
elevate potatoes and even oranges by the induction process, and it
is quite within the bounds of probability that eggs may be delivered in
this manner, without more than the usual percentage of breakages.
In conveying many materials, which are conveyed whilst hot, it is
better if they can be kept at practically the same temperature at the
delivery as when they enter the pipe. This is provided for by heating
the air to a suitable temperature just before it enters the jet.
This is also an additional source of economy in operation. As is
well known, air, like all other gases, increases in volume with the
temperature, and if the heat lost by the air cooling after compression
be replaced at the jet, considerably more power is obtained. If the
compressor is situated in such a position that most of the heat of
compression is delivered at the jet, there is little to be gained by
reheating. In most cases, however, the air has returned to normal
temperature by the time it reaches the point where it is to be used,
and, if a suitable air heater is installed at this point, the volume may
be increased greatly by a comparatively small expenditure.
A heater consisting of tubes through which the air passes, these
tubes surrounded by water under high steam pressures, offers the
most convenient and satisfactory method of heating the air. The air
pipes between the heater and conveyor pipe should be lagged in
order to retain the heat.
The pressure of the air may be increased by 50 per cent. by
heating to the temperature of steam at 200 lbs. per sq. in. gauge
pressure, while the cost will be comparatively small. Theoretically, a
gain of about 40 per cent. in economy should be obtained, and the
practical results should be reasonably close to this figure.
Air Receivers. It is a decided advantage in practice to install an
efficient separator between the ordinary receiver of the compressor
and the pipe line, as large quantities of moisture will travel over with
the air, and will be condensed directly they meet some cooler
surface. The ordinary receiver is supposed to fulfil this function, but it
does not do so because it is, in effect, an enlargement of the pipe
line, and, being filled with hot air under pressure, has no tendency to
condense the moisture. The latter does not begin to cool to any
extent until it reaches the small diameter pipes, with the
consequence that these pipes contain quantities of oil and water
which eventually reach the jet, and are blown into the material
handled.
Where compressors of the rotary or turbine type are installed,
there will be only water in expansion, but it is good practice to
remove this, even though the air be re-heated, because the moisture
will recondense in the conveyor pipe, and tend to choke the latter
when small grained substances are being conveyed.
Types of Compressors. Reference has already been made to the
power unit, and it is hardly within the scope of this work to describe
in detail the various machines available. As, however, the economy
of air conveying depends in a large measure on the cost of power, it
is evident that the compressor should be of the most suitable type for
the duty to be performed.
For small installations, single-stage reciprocating machines, driven
directly by steam engines or by electric motors are, no doubt, the
most suitable. In the case of large plants, using the air continuously
in a number of air jets, where the load factor is high, it is certainly
more important to install a two or three-stage compressor, owing to
the greater economy of working. The larger capital expenditure will
be compensated by the considerable saving of energy. As compared
with single-stage compression to 100 lbs. gauge pressure, a saving
of 20 per cent. can be effected by three-stage working, and with a
constant load of from 75 per cent. to 100 per cent. of full load, a
turbine or electrically driven rotary multiple-stage compressor is
decidedly the best type to adopt.
In plants where exhaust steam can be used to advantage, as in
large generating stations, a steam turbo-compressor, multiple-stage,
exhausting to a feed water heater will show great economy, and the
operating costs of a large plant of this type are very low compared
with any other form of conveyor. This will be obvious when it is
pointed out that maintenance costs on the conveyor are confined to
renewals of bends and junctions in the pipe lines, and of flexible
hose. There are no discharge valves or air locks to be kept vacuum
tight, no filter strainers or sleeves to renew, and the power unit is not
subjected to undue wear through extraneous matter entering the
cylinders and scoring the walls or wearing the valves.
Compared with other forms of mechanical conveying, the
pneumatic induction system is very low in maintenance costs, while
the serious charges incurred in employing human labour are reduced
to a minimum.

Full Chapter Optical Inspection of Microsystems 2Nd Edition Wolfgang Osten PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Full Chapter Optical Inspection of Microsystems 2Nd Edition Wolfgang Osten PDF

Uploaded by

Copyright:

Available Formats

Optical Inspection of Microsystems 2nd

Edition Wolfgang Osten

Optical Inspection of Microsystems First Edition

Springer Handbook of Geographic Information 2nd Edition

Elements of Optical Networking: Basics and Practice of

Essentials of Modern Optical Fiber Communication 2nd

Introduction to optical microscopy 2nd Edition Jerome

The Mexican Revolution s Wake Sarah Osten

Smart Sensors and MEMS Intelligent Devices and

The Physics of Thin Film Optical Spectra: An

© 2020 by Taylor & Francis Group, LLC

No claim to original U.S. Government works

Printed on acid-free paper

International Standard Book Number-13: 978-1-4987-7947-0 (Hardback)

and the CRC Press Web site at

Chapter 1 Image Processing and Computer Vision for MEMS Testing........................................1

Chapter 2 Surface Features.......................................................................................................... 51

Chapter 3 A Metrological Characteristics Approach to Uncertainty in Surface Metrology....... 73

Chapter 4 Image Correlation Techniques for Microsystems Inspection...................................... 93

Chapter 5 Light Scattering Techniques for the Inspection of Microcomponents and

Chapter 6 Characterization and Measurement of Microcomponents with the Atomic

Chapter 7 Optical Profiling Techniques for MEMS Measurement........................................... 177

Chapter 8 Grid and Moiré Methods for Micromeasurements...................................................207

Chapter 9 Grating (Moiré) Interferometry for In-Plane Displacement and Strain

Chapter 10 Interference Microscopy Techniques for Microsystem Characterization................ 261

Chapter 11 Measuring MEMS in Motion by Laser Doppler Vibrometry.................................. 297

Chapter 12 An Interferometric Platform for Static, Quasi-Static, and Dynamic Evaluation

Chapter 14 Digital Holography and Its Application in MEMS/MOEMS Inspection................405

Chapter 15 Speckle Metrology for Microsystem Inspection...................................................... 485

Chapter 16 Spectroscopic Techniques for MEMS Inspection.................................................... 517

Chapter 17 Sensor Fusion in Multiscale Inspection Systems..................................................... 541

• Micro-Electro-Mechanical Systems (MEMS) surface features characterization

1 China Machine Press, Beijing, China, 2014, ISBN 978-7-111-45837-1.

• Microscopic and nanoscopic measurement and testing on wafer scale

Ingrid De Wolf Markus Hüttel

Angela Duparré Xiangqian Jiang

Klaus Körner Joanna Schmit

Johann Krauter Sven Schröder

Richard Leach Georg Siegmund

1.2 CLASSIFICATION OF TASKS

1.3 IMAGE PROCESSING AND COMPUTER VISION COMPONENTS

FIGURE 1.1 Components of a computer vision scenario.

1.3.1 Behavior of Light, Colors, and Filters

FIGURE 1.2 Electromagnetic spectrum.

FIGURE 1.3 Illustration of the inverse-square law.

FIGURE 1.4 Refraction of white light by a prism.

FIGURE 1.5 Additive and subtractive colors.

FIGURE 1.6 Response of colored objects to white light.

FIGURE 1.7 Response of colored objects to colored light.

FIGURE 1.8 Effect of filter location on the resulting grayscale image.

1.3.3 Lens Systems

B is a length of flexible pipe to enable the nozzle to sweep the

Fig. 20.—Floating Pneumatic Transport Plant; 200-250 Tons

This plant is simple, but every possible modification can be added

Fig. 22.—Portable Railway Plant in Operation.

Complete plants have also been mounted on railway trucks, the

In the first case, immediately the barge is alongside, the flexible

Numerous means have been devised to cause the necessary current

You might also like