This milestone interdisciplinary work brings you to the cutting edge of emerging technologies inspired by human sight, ranging from semiconductor photoreceptors based on novel organic polymers and retinomorphic processing circuitry to low-powered devices that replicate spatial and temporal processing in the brain. Moreover, it is the first work of its kind that integrates the full range of physiological, engineering, and mathematical issues and advances together in a single source. Emphasizing both the devices and the software simulation point of view, this definitive book provides state-of-the-art retinal cell and primary visual cortex (V1) models that reflect our rapidly advancing understanding of human visual signal communication networks. It explores design and fabrication considerations behind real-world implementations, including organic light sensors that mimic human rods and cones, analog circuitry to perform retinal processing, algorithm design for motion detection and tracking, wavelet-based visual detection systems, and interest point detectors. You get the latest techniques for resolution and motion detection enhancement, including both the design and applications of biologically motivated spatio-temporal filtering of visual data, as well as a statistical framework for studying object detection in a phase-invariant manner and tools for describing local object invariants. Moreover, this trail-blazing work includes insight into the challenges that lie ahead in this cutting-edge field.
Table Of Contents
Introduction. Organic Semiconductor Photoreceptors to Mimic Human Rods and Cones Biophysics of Phototransduction in Rods and Cones. Physics of Organic Semiconductor Photosensors. Ideal Fabrication Methods. Electronic Equivalent Circuit and Electronic Models. Integration onto CMOS Arrays. Analog Retinomorphic Circuitry to Perform Retinal and Retinal Inspired Processing Foveation in the Human Eye. Retinal Processing Functions. Foveation: Electronic Versus Physical. Retinomorphic Spatial Processing. Retinomorphic Temporal Processing. Asynchronous Information Extraction. Retinal Cell Modeling Software Model. Image Processing with Irregular Sampling Patterns Techniques to Deal with Irregularly Sampled Data. Applications to Image Compression, Gradient Magnitude Extraction, and Texture Recognition. V1 Modeling Computation of a Bottom-Up Saliency Map in V1. Analog Modeling of V1 Power-Efficient Analog Signal Processing Platforms. Spatio-Temporal Filter Characterization. VLSI Design of Devices. Wavelet Models for V1 Implementing Linear Portions of VI Models. Filtering Algorithms. Use of Non-Linearities. Real-Time Implementations. Technical Challenges. From Algorithm to Hardware Implementation Design of a Novel, Biologically Inspired, Complex Steerable Wavelet Construction and Its Implementation in Reconfigurable Logic. Real-Time Saliency Maps Li 's Model. A Flexible Hardware Architecture That Can Accelerate Li 's Model. A Novel Bio-Inspired Spatiotemporal Saliency Framework. Reverse Engineering of Human Vision: Hyperacuity and Super Resolution Different Super Resolution Methods and the Reconstructions Obtained by These Methods. Motion Detection and Tracking by Mimicking Neurological Dorsal/Ventral Pathways Traditional Detection/Tracking Methods. Motion Processing in the Human Visual System. Motion Detection Using Wavelets. The Spatio-Temporal Haar Wavelet. Direction Specific Filters. Speed and Acceleration Detection. Dual-Channel Tracking Paradigm. Behavior Recognition and Understanding. ;
Anil Bharath, Ph.D. is reader in Image Analysis at Imperial College in London, where he received his Ph.D. in electrical and electronic engineering. He has published over 60 papers in the field of imaging, image analysis, and acoustics. Dr. Bharath initiated the Basic Technology Project 'Reverse Engineering Human Visual Processes', which aims to create an engineering blueprint for a subset of processes in the human visual system
Maria Petrou, Ph.D. is a professor of signal processing and the head of the Communications and Signal Processing Group at Imperial College in London. She has published more than 300 papers on topics in remote sensing, computer vision, machine learning, color analysis, medical signal and image processing, and other fields, and she is co-author of two books on image processing. Dr. Petrou is a fellow of the Royal Academy of Engineering, IEE, and IAPR, a senior member of IEEE, and a distinguished fellow of the British Machine Vision Association. She earned her Ph.D. at the Institute of Astronomy in Cambridge, UK.