Handbook of CNN Image Processing: All You Need to Know about Cellular Neural Networks (YangSky.com monographs in information sciences)

Cellular neural networks (CNN) were invented by Chua and Yang in 1988 in the Department of Electrical Engineering and Computer Sciences, University of California at Berkeley. Since then, CNN has become an extremely active field of researches to massive parallel computation, cellular automata, image processing, visual VLSI chips and vision processors. Written by one of the leading figures in the field, this is a lucid and comprehensive reference book for professionals, academic researchers and students. It covers almost all aspects of CNN including: local rules principles, structure and parameter design, continuous-time CNN, discrete-time CNN, and multi-stage CNN. Also, a systematic classification system of different CNN image operations is presented based on major local rule classes. Hundreds of CNN image operations together with their design processes were presented. The difference and equivalence between continuous-time and discrete-time CNN were formally formulated. The Generalization Principle and newly found Selected and Inverse Selected Principle make it painless to derive tens of CNN operators from a known one. 124 black-and-white figures were used to illustrate the functions of CNN image operators. Many CNN and DTCNN image operations were first time reported to the public, for example, negative CNN classes, selected CNN classes, inverse selected CNN classes, logic-morphologic CNN classes and many more. An extremely detailed index could be used as a fast and efficient looking-up table of hundreds of CNN operators.