Paper
1 October 2011 Disparity estimation by reverse fuzzyfication
Md. Abdul Mannan Mondal, Md. Haider Ali
Author Affiliations +
Proceedings Volume 8285, International Conference on Graphic and Image Processing (ICGIP 2011); 82857N (2011) https://doi.org/10.1117/12.913533
Event: 2011 International Conference on Graphic and Image Processing, 2011, Cairo, Egypt
Abstract
This paper presents an algorithm using fuzzy logic for computing disparity or stereo correspondence of the image sequences. For better achievements, reverse fuzzyfication method has been employed for disparity estimation. The algorithm stands on verdict the preeminent counterpart for every window's pixel in the left image with the consequent right image of all the pixels and setting the fuzzy value for measuring the stereo correspondence. In this method 3×3 window size has been used and every pixel has been considered for estimating stereo correspondence. Discrete membership function is used based on the basis of the pixel value difference between left and right images for fuzzyfication. Experimental results demonstrate that the proposed algorithm's computational cost is much less than the fixed and adaptive window based methods, and it requires less memory space too. The proposed method has been applied to standard stereo images and the results imply that we can easily reduce the computational time of about 50% with no appreciable degradation of accuracy.
© (2011) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Md. Abdul Mannan Mondal and Md. Haider Ali "Disparity estimation by reverse fuzzyfication", Proc. SPIE 8285, International Conference on Graphic and Image Processing (ICGIP 2011), 82857N (1 October 2011); https://doi.org/10.1117/12.913533
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Fuzzy logic

Cameras

Image processing

Fuzzy systems

Image compression

Image quality

Machine vision

Back to Top