A New system for offline Signature Identification and verification

BIOMETRIC FEATURES HAVE GREAT IMPORTANCE IN AUTHENTICATION SYSTEMS NOWADAYS. ONE OF THE MOST IMPORTANT AND CONVENTIONAL BIOMETRICS IS SIGNATURE. IN THIS PAPER, WE PROPOSED A SYSTEM WHICH HAS TWO INDEPENDENT PHASES FOR OFFLINE SIGNATURE IDENTIFICATION AND VERIFICATION. THE IDENTIFICATION PHASE IS BASED ON TSR (TRIANGLE SPATIAL RELATIONSHIP) THAT IS A ROTATION INVARIANT FEATURE EXTRACTION METHOD. ALSO, A SYMBOLIC REPRESENTATION OF SIGNATURE HAS BEEN EMPLOYED TO MAKE USING TSR POSSIBLE. IN THE VERIFICATION PHASE, A HYBRID METHOD IS PROPOSED THAT COMBINES DISCRETE WAVELET TRANSFORM, GABOR FILTER, AND IMAGE FUSION METHODS. EXPERIMENTAL RESULTS ON SOME BENCHMARKS HAVE CONFIRMED THE ROBUSTNESS AND PRECISION OF PROPOSED METHOD TOGETHER WITH ITS ROBUSTNESS AGAINST TRANSLATION, SCALING, AND ROTATION.