Cellular Neural Networks and Image Processing
This book applies the design principles of cellular image operators to a hardware platform called cellular neural network (CNN). CNN is a member of the hardware family called vision chips. Based on state-of-the-art technology, a vision chip is defined as a VLSI chip that can perform image processing tasks. CNN is mostly nourished from two main fields: One is cellular automata and the other is neural network. As an interdisciplinary product, the study of CNN mainly focuses on finding specialised design principle called template design. CNN utilises cellular hardware structures to gain ultrahigh image processing speed.
What people are saying - Write a review
We haven't found any reviews in the usual places.
Design Principles of CNN
CNN Implementations of Local Rule Classes
5 other sections not shown
arbitrary boundary condition BBB BBB black cells pixels CNN implementation CNN structure CNN This CNN contradiction DBB DBB DBB DBB DDB DDB DDB DDB DDD DDD DTCNN edge CNN edge_i edge-detecting CNN feasible parameter range feasible parameter regions final output fixed 1 boundary following Eq following three-stage CNN follows from Eq follows from Theorem image at stage image processing implementation is given Implementing local rule Implementing logic difference initial condition Initial state insensitive input image insensitive implementation aoo inseparable local rules Local Rule Class logic difference CNN logic difference operation MCNN namely number of black original local rule output image output is black output is white p=0 aoo parameter 0,0 phase plane R-point RC-4C edge rule bifurcation rule with parameter rules are given shown in Fig simulation result source image stage is f/'1 third inequalities three-stage CNN algorithm uncoupled CNN yW(oo yy(oo