TERM PAPER

BIOINFORMATICS

TOPIC

Writer Identification with Hybrid Edge Directional Features and Nonlinear Dimension Reduction

WRITER IDENTIFICATION

Writer identification belong to the group of behavioral methods in biometrics. Contrary to biometrics with a purely physical or biophysical basis, the bio-metric analysis of handwriting requires a very broad knowledge at multiple levels of observation. The significance and scope of writer identification is becoming more prominent these days.

 A good biometric is universal, unique, permanent and collectable. UNIVERSALITY Handwriting is not fully universal, since there exists still a considerable proportion of non-writing individuals in the population. UNIQUENESS Uniqueness can be only determined empirically: assuming a natural writing attitude, there is more individual information in handwriting than is generally assumed, especially if one has enough text per sample

PERMANANCE Handwriting does change over time: both for signature verification and hand-writing identification and verification, regular enrollment procedures over the years will be required COLLECTABILITY Samples can be collected on-line, using a special digitizer which samples the pen-tip position and possibly other movement-related signals in time. Alternatively, existing samples of handwriting (offline) can be scanned.

 The writer identification techniques is currently used in wide arenas like the digital rights administration, forensic expert decision-making systems, document analysis systems and also as a strong tool for physiological identification purposes.

The writer identification is also utilized in authentication system by combining with the writer verification for many fields of confidential data handling purposes.

EDGE DIRECTIONAL FEATURES

Feature extraction starts with conventional edge detection that generates a binary image in which only the edge pixels are on.

 We then consider each edge pixel in the middle of a square neighborhood and we check (using logical AND operator) in all directions emerging from the central pixel and ending on the periphery of the neighborhood for the presence of an entire edge fragment (see fig. 1). All the verified instances are counted into a histogram that is finally normalized to a probability distribution which gives the probability of finding in the image an edge fragment oriented at the angle measured from the horizontal. In order to avoid redundancy , the algorithm only checks the upper two quadrants in the neighborhood because, with-out on-line information, we do not know which way the writer traveled along the found oriented edge fragment

NON-LINEAR DIMENSION REDUCTION

Reducing dimensionality of data with minimal information loss is important for feature extraction, compact coding and computational efficiency. The data can be transformed into good representations for further processing, constraints among feature variables may be identified, and redundancy eliminated.

THANK YOU