Synopses & Reviews
Depth recovery is important in machine vision applications when a 3-dimensional structure must be derived from 2-dimensional images. This is an active area of research with applications ranging from industrial robotics to military imaging. This book provides the comprehensive details of the methodology, along with the complete mathematics and algorithms involved. Many new models, both deterministic and statistical, are introduced.
Synopsis
Computer vision is becoming increasingly important in several industrial applications such as automated inspection, robotic manipulations and autonomous vehicle guidance. These tasks are performed in a 3-D world and it is imperative to gather reliable information on the 3-D structure of the scene. This book is about passive techniques for depth recovery, where the scene is illuminated only by natural light as opposed to active methods where a special lighting device is used for scene illumination. Passive methods have a wider range of applicability and also correspond to the way humans infer 3-D structure from visual images.
Table of Contents
Passive Methods for Depth Recovery.- Depth Recovery from Defocused Images.- Mathematical Background.- Depth Recovery with a Block Shift-Variant Blur Model.- Space-Variant Filtering Models for Recovering Depth.- An ML Estimator of Depth and Optimal Camera Settings.- Recursive Computation of Depth from Multiple Images.- MRF Model-Based Identification of Shift-Variant PSF.- Simultaneous Recovery of Depth and Image Restoration.- Conclusions.