The application of neural networks to image processing and vision tasks was very popular. Originally published in 1995, this book dealt also with images and neural networks, but with a different goal. The primary focus was not on the application of neural networks to images but on the transfer of knowledge between image processing/computer vision and neural networks. In order to cope with the complexity inherent in most vision tasks, divide-and-conquer techniques had to be applied that naturally led to hierarchical structures. It is very important to exploit the capabilities of hierarchies in a systematic way. The approach taken in this book was by considering the similarity between hierarchical neural networks and image pyramids. Throughout this book, all aspects of this similarity are considered, and various results are presented that demonstrate the advantage of this approach. The neural networks resulting from this study have many similarities to image pyramids and are therefore called pyramidal neural networks.