INVARIANT OPTICAL-PATTERN RECOGNITION BASED ON A CONTOUR BANK

We propose a new optical pattern-recognition method that is invariant to shift, rotation, and scale changes. By means of a zero-order circul ar-harmonic filter the input object can be transformed into a correlat ion peak whose outline a shape similar to the outline of the input obj ect. After multilevel thresholding and binarization, a bank composed o f many contours with different sizes is formed. The invariant recognit ion can be obtained by a small contour line being taking as a referenc e and its second-order circular harmonic component being taken as a fi lter. In addition to the scale invariance the correlation peak is supe rior to the classical circular-harmonic filter and the phase-only circ ular-harmonic filter in terms of peak intensity, half-maximum width, a nd peak-to-sidelobe ratio.