Determination of egomotion and environmental layout from noisy time-varying image velocity in binocular image sequences

Authors:
John L. Barron;Allan D. Jepson;John K. Tsotsos
Affiliations:
Department of Computer Science, University of Toronto, Toronto, Canada;Alio, Canadian Institute for Advanced Research and Department of Computer Science, University of Toronto, Toronto, Canada;Alio, Canadian Institute for Advanced Research and Department of Computer Science, University of Toronto, Toronto, Canada
Venue:
IJCAI'87 Proceedings of the 10th international joint conference on Artificial intelligence - Volume 2
Year:
1987

Citing 4
Cited 3

Binocular Image Flows: Steps Toward Stereo-Motion Fusion

IEEE Transactions on Pattern Analysis and Machine Intelligence
Computational Stereo

ACM Computing Surveys (CSUR)
Determinations of egomotion and environmental layout from noisy time varying image velocity information in monocular image sequences

Determinations of egomotion and environmental layout from noisy time varying image velocity information in monocular image sequences
Visually Interpreting the Motion of Objects in Space

Computer

Binocular estimation of motion and structure from long sequences using optical flow without correspondence

ICIP '95 Proceedings of the 1995 International Conference on Image Processing (Vol.2)-Volume 2 - Volume 2
The sensitivity of motion and structure computations

AAAI'87 Proceedings of the sixth National conference on Artificial intelligence - Volume 2
The sensitivity of motion and structure computations

AAAI'87 Proceedings of the sixth National conference on Artificial intelligence - Volume 2

Quantified Score

Hi-index	0.00

Visualization

Abstract

We address the problem of interpreting time-varying image velocity fields generated by a moving binocular observer viewing a stationary environment under perspective projection to obtain 3-D information about the absolute motion of the observer (egomotion) and the absolute depth of environmental surface points. We conduct a numerical study of our algorithm (which involves solving non-linear systems of equations) for best, random and worst case image velocity error. As well, we investigate how good the initial guess for the nonlinear system of equations has to be. Other results include the presence of multiple solutions in time, how the algorithm performs when the underlying assumptions it is based on are violated and the effect of varying the spatial extent of the image points used, of varying the spatial baseline (separation of the left and right cameras) and of varying the temporal extents of the image points used (effectively varying the temporal baselines). As well, we investigate the use of convergent/divergent as opposed to parallel stereo camera setups.