Pulse-coupled neural networks (PCNNs) are based on the nervous system in the eyes of humans and mammals. The book describes various models of PCNNs, and related image processing algorithms. Image processing techniques such as image filtering, image segmentation, image coding, image enhancement, image fusion, and feature extraction are covered. The book also has a chapter on the application of PCNNs to optimization problems, and one on a hardware implementation of PCNNs.

Chapter 1 introduces PCNNs and their basic mathematical formulas, which relate to properties of PCNNs, such as impulse response, linking input, feeding input, and the firing frequency of neurons. It explains the mechanisms of a PCNN through diagrams. It also presents the simplified intersection cortical model, the spiking cortical model, and the multichannel PCNN.

Chapter 2 presents three algorithms for image filtering. An image filtering algorithm takes an unclear image and outputs a clear image. This chapter explains filters such as the mean filter, the median filter, the morphological filter, and the minimum mean square filter. For each algorithm, the experimental results are discussed along with the presentation of its flow chart or procedure. Actual image results are shown.

In chapter 3, the authors describe image segmentation using PCNNs and genetic algorithms (GAs). The description includes the PCNN model used, the application scheme for GAs, the algorithm flow, and the experimental results. The chapter also presents image segmentation using PCNNs and entropy. Traditional image segmentation techniques are compared with the PCNN entropy technique.

Chapter 4 describes an algorithm for image compression and coding. Irregular segmented region coding (ISRC) based on PCNNs is explained, along with the framework, experimental results, and theories behind ISRC, including orthogonality and chain code. The analysis of experimental results features a comparison with classical image compression algorithms using parameters such as compression ratio and quality of the reconstructed image.

Chapter 5 is on image enhancement, in which an image is processed to improve visual effects for a specific purpose. It presents image enhancement using a PCNN time matrix. The PCNN time matrix is described in detail. The chapter first introduces various image enhancement methods such as gray-level transformation, histogram processing, and filtering. The relation between PCNNs and human visual characteristics is discussed, followed by a description of the PCNN time matrix.

Image fusion takes many images and produces a more reliable fused image. Chapter 6 describes an image fusion algorithm for m-PCNN, a variant of PCNN. This is followed by an algorithm for multi-focus image fusion on a dual-channel PCNN, which is n-PCNN with n as 2. Multi-focus image fusion fuses images of the same object with different focus settings.

Chapter 7 discusses pattern recognition using PCNNs, and introduces various transformations required for feature extraction from images using PCNNs. Texture feature extraction is described, along with a detailed analysis of experimental results. An iris recognition system is described, including the various steps involved, the PCNN-related formulas, and the experimental results.

Chapter 8 is on solving combinatorial optimization problems using PCNNs. After introducing such problems, the chapter presents one method to solve the shortest path problem, and another to solve the traveling salesman problem. Two new PCNN variants are introduced: an auto-wave neural network and a tristate cascading PCNN.

In chapter 9, the authors consider hardware implementation of a PCNN algorithm on a field-programmable gate array (FPGA) system. The description includes block diagrams of the implementation, and an analysis of experimental results for various images.

This book could serve as a textbook on PCNNs. Someone who is generally acquainted with image processing should be able to understand the image processing algorithms in the book. Each algorithm features not only the flow of the algorithm, but also a detailed empirical analysis. The book has a very large number of references for researchers. A background in elementary statistical notation is required to understand the formulas in the book.