Brains are characterized by limited bandwidth and computational resources. At any point in time, we can focus our attention only to a limited set of features or objects. One of the most remarkable –and often neglected– differences between machine vision and biological vision is that computers are often asked to process an entire image in one shot and produce an immediate answer whereas animals ...

Foveated vision and two-mode tracking, as inspired b y the human oculomotor system, are often used in active vision system. The purpose of this paper is t o provide answers to the following basic questions which arise from implementations. First, is it beneficial to have foveated vision and what is the optimal size of the foveal window? Second, is there a need for two control mechanisms (smooth...

A number of recent papers have demonstrated that camera "selfcalibration" can be accomplished purely from image measurements, without requiring special calibration objects or known camera motion. We describe a method, based on self-calibration, for obtaining (scaled) Euclidean structure from multiple uncalibrated perspective images using only point matches between views. The method is in two st...

This paper addresses four questions concerning control of attention: how to direct attention to novel or ‘interesting’ events in the world, how to separate novel or interesting information from everything else, how to perform this segmentation of the world at multiple levels, how to develop perceptual machinery capable of this. To all four questions we provide answers arising from work in our W...

Foveal or spatially-variant image representations are important components of active vision systems. Log-polar sampling is a particularly powerful example as a result of the simplicity with which expansion and rotation can be handled. These properties are exploited here for the detection of general straight lines, line segments, and circles through the foveation point. An efficient and practica...

This paper addresses a classical problem-optical ow-for which we propose novel estimation algorithms based on robust regression. Another important aspect of our work is that we are concerned with the computational eeciency of our estimators, in the sense of maximizing its performance, for a chosen computation cost. This concern about eecient estimation appears often in active vision. It poses a...

The Gabor decomposition is a ubiquitous tool in computer vision. Nevertheless, it is generally considered computationally demanding for active vision applications. We suggest an attention{driven approach to feature detection inspired by the human saccadic system. A dramatic speedup is achieved by computing the Gabor decomposition only on the points of a sparse retinotopic grid. An application t...

Introduction Brains and sensory systems are characterized by limited bandwidth and computational resources. At any point in time, we can focus our attention only to a limited set of features or objects. One of the most remarkable –and often neglected– differences between machine vision and biological vision is that computers are often asked to process an entire image in one shot and produce an ...

Probably the most ambitious goal of Computer Vision is to build the universal vision machine, capable of guiding itself through arbitrary environments, recognizing objects along its path and reaching its destination, wherever that might be. In spite of this elusive goal, Computer Vision is a eld of research where enormous progress has been accomplished. In particular, the paradigm of Active Vis...

This article discusses the development of a mobile bus-mounted machine vision system for transit and traffic monitoring in urban corridors, as required by intelligent transportation systems. In contrast to earlier machine vision technologies used for traffic management, which rely mainly on fixed-point detection and simpler algorithms to detect certain traffic characteristics, the new proposed ...