Abstract

Recent developments in signal processing, vision, and image understanding reveal that a proper signal or image decomposition before the actual processing may provide enormously useful information about the signal or image. The wavelet transform a particularly useful model, provides a multi resolution linear signal/image time-frequency or space-frequency decomposition tool. Successful applications of wavelet transforms to solve difficult signal or image analysis and synthesis problems have been widely reported. Digital implementations of these transforms are computationally intensive both because of the nature of the coordinate doubling of the wavelet transform and because of the large quantity of convolution/correlation operations that accompany them. Optics with its inherent parallel processing capability, has been used to obtain many useful linear signal/image transformations and will certainly have some roles in wavelet-based processing. This survey talk is intended to outline the suitability of using existing optical processing techniques for wavelet-based signal/image analysis and synthesis. A brief summary of various proposed optical wavelet transform approaches together with their advantages and constraints will be presented. Possible future directions in this new research field will also be discussed.

© 1992 Optical Society of America