GsE 189 Lecture 13 Lidar RS

Department of Geodetic Engineering
4/21/2016
Training Center for Applied Geodesy and Photogrammetry GsE 189 – Remote Sensing: Theory and Applications
1
4/21/2016
2
INTRODUCTION TO
LIDAR REMOTE SENSING
LECTURE 13
GsE 189 - REMOTE SENSING: THEORY AND APPLICATIONS
Lecturer: Ayin M. Tamondong

4/21/2016
3
Traditional Methods of Terrain Data
Collection
• Conventional ground
surveys (transits and
theodolites)
• Electronic Distance
Measurement (EDM)
Devices
• Global Positioning
Systems
• Photogrammetric
Mapping
4/21/2016
4
Traditional Methods of Terrain Data
Collection
• Problems with these methods
• Labor Intensive
• Time-consuming
• Costly
• Dictated by conditions (time of year,
sun angle, weather, etc.)

4/21/2016
5
What is LiDAR?
Light
Detection
And
Ranging
Analogous to RADAR, but using the infra-red
(also UV, visible) portion of the EM spectrum

4/21/2016
6
Lidar System Overview
• Measures distances (through laser pulses) that
strike and reflect from features on the surface of
the earth.
• Distance = (speed of light * round-trip time)* ½
• LiDAR is the application of a number of modern
technologies: Positioning, Inertial technology, Laser
scanning Digital imagery
• to provide fast, efficient and cost-effective
surveys using established survey principles
• Converts scanning angle and distance-from-sensor
information into georeferenced data points: 100K-
500K positions/second (point clouds)

4/21/2016
7
Airborne Lidar System Components
GPS IMU
z
pitch
• Scanning laser emitter-receiver
y
unit
X roll
yaw • Differentially-corrected GPS on
the ground and in the plane
• Inertial measurement unit (IMU)
• Computer to control the system
GPS and store data
z
y
X

4/21/2016
8
LiDAR Principle
Velocity z Operating altitude:
y 400 - 2,000 m.
x
x
If we know
y Position
z x, y & z
Pitch, roll & Angle
yaw  Range
r We can derive
Position
x, y & z
Laser pulse rate: up

to 25,000 per sec.
Swath width: up to 1,500 m. at 2,000 m. altitude.

4/21/2016
9
How it All Works
GPS Satellites
GPS Receivers
Base Station

4/21/2016
10
LiDAR Advantages
• Provides a highly accurate means of elevation model
collection for 0.3 or 0.6m contours
• Acquisition can take place day or night… shadows
that are problematic in mountainous areas are not
an issue with LiDAR
• Acquisition can take place below cloud cover… cloud
shadows no issue
• Very cost effective for larger projects
• LiDAR performs better in vegetated areas than

photogrammetry
• Able to collect data in “leaf on” condition
4/21/2016
11
LiDAR Advantages
• LiDAR can offer considerable advantages over
conventional survey or photogrammetry for
the right project
• Fast data collection
• Fast data processing (dependent on deliverables)
• Little or no need for access (in office data mining)
• Less weather dependent than photogrammetry
• Night operations
• Not if imagery is required
• Robust data sets with many possible products
• Cost savings (economies of scale)

4/21/2016
12
Why is This Technology Exciting
Time to Collect 1 Million Points

• Conventional Surveying: 15.5 years
• Photogrammetry: 1.5 years
• Lidar: 6.7 seconds @ 150 kHz
Costs can be significantly less for the

right projects…

4/21/2016
13
Issues with LIDAR Data
• LIDAR is indiscriminate… it places elevation

points on everything. This includes cars,
houses, trees, etc.
• LIDAR only places mass points, or random

xyz points. It does NOT pick up breaklines,
or lines of abrupt change in the ground
elevation
• LIDAR is NOT imagery. LIDAR data can be

shaded, however, to offer a relief image
4/21/2016
14
Issues with LIDAR Data
• LIDAR is not capable of penetrating thick
vegetation
• Supplemental information cannot be derived

from LIDAR
• It is not an all weather system
• LiDAR cannot identify the following features:

• Boundary information
• Underground utilities

4/21/2016
15
LiDAR based on platform
• Space-borne LiDAR
• Airborne LiDAR
• Terrestrial LiDAR
• Terrestrial Laser Scanner

4/21/2016
16
Types of Airborne LiDAR
• Topographic LiDAR
Wide Area
Systems
Corridor
Mapping
• Bathymetric LiDAR Systems

4/21/2016
17
Lidar Data Characteristics
• A lidar “Pulse” is a laser beam
emitted from the Lidar sensor
(NIR –1064nm, G –532nm).
• Scan rate ~ up to 500 kHZ (500,000
pulses per second)
• A single pulse may strike several
features
• Portions of the pulse are reflected
back to the sensor, these portions
are called “returns”
• Up to 5 returns per pulse
• *Typically only 2-3 returns
1st Return (Canopy Ht)
Intermediate Return
Intermediate Return
Last Return (Ground)
4/21/2016
18
• Data Resolution (Full Point Cloud)
• Pulses per unit area or Pulse Density (pulses/m2). Important
when vegetation derivatives are required from the lidar data.
• Return density can vary depending on the target being
scanned.
1st Return (Canopy Ht)

Intermediate Return
Intermediate Return
Last Return (Ground)

4/21/2016
19
• Lidar Intensity - each return is
associated with an intensity
value 1m pixel
• Relative measure of the
reflectivity of the target (1064
nm wavelength most common)
• Consistency
• between scanners or missions
• between returns in the same
pulse
• Sensor may vary its sensitivity
(gain) during the flight to
compensate for variability in
return signal strength

4/21/2016
20
• Discrete-return Lidar
data consists of
precisely referenced
points in time and space.
• A collection of these
points are referred to as
a “point cloud”. From
point clouds we can
extract information
about:
• Ground
• Roads and other surface
features
• Vegetation
• Structures
4/21/2016
21
Intensity and Multiple Returns
• Most Lidar units today have the ability to
measure multiple returns and the
intensity of the returned signal for each
• This enables specialized applications
using the LIDAR data

4/21/2016
22
Return Density
LiDAR sensor systems vary in the

number of returns from a surface

4/21/2016
23
Multiple LIDAR Returns

4/21/2016
24
Multiple Returns
Full
stnd Filtered
3rdReturns
14th
2
VegetatedReturns
Returns
GroundROW Set

4/21/2016
25
What Information can Lidar Provide?
• High Resolution (1
meter) “Bare Earth
Surface” or DEM
• Roads, Structures and

other Surface features
• Vegetation (canopy
closure, tree and
stand heights,
canopy Structure,
etc…)
4/21/2016
26
Bare Earth Point Clouds
• Filtered to contain only points classified as ground
returns (vegetation, structures, etc. filtered out)
Bare Earth Filtered Point Cloud
When assessing a bare
earth filtered point cloud,
Point Spacing indicates
how well we sampled the
“ground”
Raw Point Cloud – all features included
When assessing a raw lidar

point cloud, Pulse Density
indicates how well we
sampled “all targets”

4/21/2016
27
Deliverables
• A LiDAR “point cloud” looks impressive,
but is of little use to most clients
• Accuracy, detail and deliverable
requirements will dictate which type of
LiDAR system should be used for data
collection
• Raw LiDAR data sets are very large
• Filtering
• Digitizing
• Vectorizing
…can reduce data sets to be manageable,
while maintaining accuracy and detail

4/21/2016
28
4/21/2016
29
Jet

4/21/2016
30
470’ Tall
Mississippi River
4/21/2016
31
4/21/2016
32
LiDAR Returns and Cloud Cover

4/21/2016
33
Cincinnati Airport – Aerial Photo

4/21/2016
34
Cincinnati Airport – Intensity Plot

4/21/2016
35
Filtering: Effect of Vegetation Removal
Before After
4/21/2016
36
Bare Earth Surface or DEM

4/21/2016
37
Highest Hit Surface or “Canopy Surface Model”

4/21/2016
38
Derived Lidar Products
Canopy Surface Model
Bare Earth Surface

4/21/2016
39
Derived Lidar Products
• By using the ground surface we can normalize the
topography or remove it from the data.
• This allows us to measure and make comparisons of
vegetation structure (Canopy Metrics)

4/21/2016
40
Vegetation Derivatives

4/21/2016
41
Return data colored by height
Individual Tree
Metrics
Plot or Grid “Stand” Metrics

Metrics (return heights)
Min, Max, Mean
Median, Mode, Std Dev
Percentile heights
Skewness, Kurtosis
Cover above 2m
Cover above mean
Cover above mode
Same metrics using return intensity values

4/21/2016
42
Lidar Pulse Density and Products
• Low Pulse Density (≤ 1 Pulse/m2)

• Product: Moderate Resolution Topographic
Products (≥ 2 meter Grid)
$$
• Moderate Pulse Density (1-3 Pulses/m2)

• Products: Stand Level Vegetation Metrics (e.g.
canopy height, canopy cover) and High
Resolution Topographic Products
$$$
• High Pulse Density (≥ 3 Pulses/m2)

• Products: Forest Structure
• Modeling and Field Data Collection required
$$$$
4/21/2016 GsE 189 – Remote Sensing: Theory and Applications
43
Training Center for Applied Geodesy and Photogrammetry
Low Pulse Density (≤ 1 Pulse/m2)
• Moderate Resolution (3m) Lidar Dem can provide:

• Improved Stream Location Accuracy
• Improved Flood Mapping
• Landslide Detection and hazard assessment
4/21/2016
44
Moderate Pulse Density (1-3 Pulses/m2)
• Stand Level Vegetation Metrics
Legend
Mature Loblolly Stand

• 85 years old
• Area = 65 acres
Young Loblolly Stand

• 19 years old
• Area = 62 acres
Canopy
% Canopy
Height
Cover(m)
High:40m
High:100%
Low:
Low:0%
0m
 Canopy
% Canopy height
covermodel
model
created from the Sabine NF
NF lidar
lidar data.
data.
Background image represents a subset (1150 acres) of the Sabine National Forest, TX. (NAIP
08, 1 meter resolution)
4/21/2016
45
High Pulse Density (≥ 3 Pulses/m2)
• Includes all capabilities of Low and Mod. Pulse density data
• Possible to model Forest Inventory parameters: Dominant height,
basal area, stem volume, biomass, canopy fuel variables, stem density
• Expensive: lidar data, Accurate field data, complex analysis
Legend
Build Inventory Models
Collect Data Basal Area (SqFt/Acre)
Explore Relationship High: 563
Low: 3
Lidar Plot Metrics
• Build a statistical
relationship between the
lidar plot metrics and field • Use the statistical relationship to
plots measurements. predict Forestry Inventory
Field Plot Measurements parameters across the entire lidar
acquisition.

4/21/2016
46
High Pulse Density (≥ 3 Pulses/m2): Model Outputs
Biomass Canopy Fuel Load
R²=.88 R²=.6
Total Basal Area Total Volume
R²=.78
Department of Geodetic Engineering R²=.82
4/21/2016
47
LIDAR Errors
• Laser induced – changes in height for points

on terrain (ridges and ditches) and grain
noise (smooth surfaces appear rough)
• GPS/INS induced – variances in
measurements taken by the instruments
• Filtering induced – incomplete/unnecessary
removal of features (vegetation, buildings,
rock outcroppings)

4/21/2016
48
4/21/2016
49
4/21/2016
50
Error Sources
• Aerosols and Dust

• Scatter Laser reducing signal strength of Laser
reaching target
• Laser reflected to receiver off of dust introduces
noise
• Minimally sensitive to temperature variation
(changes path length inside of receiver and
clock oscillator rate)
• Error in measurement of rotating mirror
angle
• Specular Surfaces
• Clock Error
4/21/2016
51
LiDAR Accuracy
• Quality of the hardware and software
• Knowledge of the planners, operators, office
staff
• Flying height
• Scan angle (also important for vegetation
penetration)
• GPS configuration
• Distance from base station to aerial platform
• Laser power
• Laser rep rate
4/21/2016
52
LIDAR Accuracy
• Accuracy of elevation in range of 6 to 30

centimeters
• Accuracy of XY position in range of 10 to 46
centimeters
• Accuracy depends on pulse rate, flying height,
GPS configuration, location of ground
stations, and position of the scanner with
respect to nadir

4/21/2016
53
Additional Complementary Sensors
• LiDAR systems are often fitted with additional
sensors which can provide useful information to
accompany the LiDAR data
• Digital Imagery -- Video/Still
• Orthophotography
• Infra-red cameras
• Hyperspectal sensors

4/21/2016
54
Common LiDAR Applications
• Base mapping
• Floodplain mapping
• Natural resources management
• Transportation and utility mapping
• Urban modeling
• Transmission lines,
• Railways, highways
• Levees, pipelines

4/21/2016
56
Sources
• Fundamentals of RS by CCRS, pp 92.

• Introduction to Lidar Technology and FUSION Software by
Brent Mitchell (USDA Forest Service).
• Introduction to Lidar Remote Sensing Lecture Notes (GE 202)
by E. Paringit.
• Remote Sensing by R. Schowengerdt, pp. 10.
• RS Digital Image Analysis by J. Richard & X. Jia, pp 407.

4/21/2016
58 58
Thank you for your attention 

4/21/2016
59 59

GsE 189 Lecture 13 Lidar RS

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

GsE 189 Lecture 13 Lidar RS

Uploaded by

Copyright:

Available Formats

Department of Geodetic Engineering

Lecturer: Ayin M. Tamondong

Department of Geodetic Engineering

Department of Geodetic Engineering

Department of Geodetic Engineering

Department of Geodetic Engineering

Laser pulse rate: up

Department of Geodetic Engineering

Department of Geodetic Engineering

• LiDAR performs better in vegetated areas than

Department of Geodetic Engineering

Time to Collect 1 Million Points

Costs can be significantly less for the

Department of Geodetic Engineering

• LIDAR is indiscriminate… it places elevation

• LIDAR only places mass points, or random

• LIDAR is NOT imagery. LIDAR data can be

• Supplemental information cannot be derived

• It is not an all weather system

• LiDAR cannot identify the following features:

Department of Geodetic Engineering

Department of Geodetic Engineering

Department of Geodetic Engineering

1st Return (Canopy Ht)

Department of Geodetic Engineering

Department of Geodetic Engineering

Department of Geodetic Engineering

LiDAR sensor systems vary in the

Department of Geodetic Engineering

Department of Geodetic Engineering

Department of Geodetic Engineering

• Roads, Structures and

When assessing a raw lidar

Department of Geodetic Engineering

Department of Geodetic Engineering

Department of Geodetic Engineering

Department of Geodetic Engineering

Department of Geodetic Engineering

Department of Geodetic Engineering

Department of Geodetic Engineering

Department of Geodetic Engineering

Bare Earth Surface

Department of Geodetic Engineering

Department of Geodetic Engineering

Department of Geodetic Engineering

Plot or Grid “Stand” Metrics

Department of Geodetic Engineering

• Low Pulse Density (≤ 1 Pulse/m2)

• Moderate Pulse Density (1-3 Pulses/m2)

• High Pulse Density (≥ 3 Pulses/m2)

• Moderate Resolution (3m) Lidar Dem can provide:

Mature Loblolly Stand

Young Loblolly Stand

Lidar Plot Metrics

Department of Geodetic Engineering

• Laser induced – changes in height for points

Department of Geodetic Engineering

• Aerosols and Dust

• Accuracy of elevation in range of 6 to 30

Department of Geodetic Engineering

Department of Geodetic Engineering

Department of Geodetic Engineering

• Fundamentals of RS by CCRS, pp 92.