Stereo Correspondence Using Color Based on Competitive-cooperative Neural Networks

Authors:
Xijun Hua;Masahiro Yokomichi;Michio Kono
Affiliations:
Jiangsu University, Zhenjiang, China;University of Miyazaki, Miyazaki, Japan;University of Miyazaki, Miyazaki, Japan
Venue:
PDCAT '05 Proceedings of the Sixth International Conference on Parallel and Distributed Computing Applications and Technologies
Year:
2005

Citing 0
Cited 1

Biologically and psychophysically inspired adaptive support weights algorithm for stereo correspondence

Robotics and Autonomous Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

In the literature, most of the stereo matching methods have been limited to gray level images, only few authors have dealt with color images straightly. In this paper, we propose a novel area-based color stereo matching method based on competitive-cooperative neural networks. Seven kinds of color spaces are tested in order to evaluate their suitability to stereo matching. The experimental results show that the matching precision is increased efficiently, when using adaptive color features instead of gray values.According to the experimental results, Ohta, Opponent and YCbCr color spaces can offer good color features for stereo matching.