Vision - Robotics Institute Carnegie Mellon University

Vision

Book Section/Chapter, Handbook of Artificial Intelligence, Vol. 3, pp. 125 - 321, 1982

Abstract

Vision is the information-processing task of understanding a scene from its projected images. An image is a two-dimensional function f(x, y), obtained with a sensing device that records the value of an image feature at all points (x, y). Values might be binary for black-or-white images, gray level for half-tone images, or vectors of color measures for color images. Images are converted into a digital form for processing with a computer. An array {fi,j} of small picture-elements called pixels represents the image by recording the values of measurements at each pixel position. The task of a computer-vision system is to understand the scene that an image—an array of pixels—depicts. This chapter discusses image-understanding research. There are a few levels of information processing in computer vision. Low-level vision or early processing systems extract primitive features, such as change of intensity and orientation of edge elements, from the original intensity array.

BibTeX

@incollection{Kanade-1982-15602,
author = {Takeo Kanade},
title = {Vision},
booktitle = {Handbook of Artificial Intelligence},
publisher = {W. Kaufmann, Inc.},
chapter = {13},
editor = {P. Cohen and E. Feigenbaum},
year = {1982},
month = {January},
pages = {125 - 321},
volume = {3},
}