Recognition of 2-D Optical Pattern by using Simulated Annealing Algorithm
Dr. Pawan Kumar
Associate Prof., Department of Physics, A.S. (P.G.) College, Sikandrabad, India
A classical optical correlator cannot give better performance in the recognition process when the input has different scale and orientation with respect to reference image. Correlation performance can be improved using techniques such as Mellin transforms for scale invariance, circular harmonic expansion for rotation-invariance and synthetic discriminant function for scale and rotation invariances. In the present study, a bipolar composite filter (BCF) is design using simulated annealing algorithm. It can be a rotation-invariant or scale invariant or both depending on one’s choice. In the present study, it is constructed for out of scale pattern recognition it is noticed that BCF offers high discrimination capability for targets and anti-target patterns.
Keywords: Correlation, Bipolar Composite Filter (BCF), Pattern recognition.