Improved Disparity Map Estimation from Multiple Images on Hybrid Method
Chhatrala Nayankumar D1. Bhalodiya Kelvin2 J. Doshi Kaushal J.3
1Chhatrala Nayankumar D., Dept. of Electronics & Communication Engineering, Marwadi Education Foundation Group Of Institution, Rajkot, Gujarat, India.
2Bhalodiya Kelvin, Dept. of Electronics & Communication Engineering, Marwadi Education Foundation Group Of Institution, Rajkot, Gujarat, India.
3Doshi Kaushal J., Dept. of Electronics & Communication Engineering, Marwadi Education Foundation Group Of Institution, Rajkot, Gujarat, India.
Manuscript received on January 25, 2014. | Revised Manuscript received on February 17, 2014. | Manuscript published on February 28, 2014. | PP: 115-117 | Volume-3, Issue-3, February 2014. | Retrieval Number: C2599023314/2013©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Stereo vision systems aim at reconstructing 3D scenes by matching two or more images taken from slightly different viewpoints. The main problem that has to be solved is the identification of corresponding pixels, i.e. pixels that represent the same point in the scene In this paper, a stereo matching algorithm based on image segmentation is presented. We propose the hybrid algorithm based on k-means segmentation and refine the disparity map of the stereo image by SSD (sum of squared difference).Firstly, a color based k-means segmentation method is applied for segmenting the stereo images. segmentation is used to represent, locate and analyze the image. Segmented images are used as input to the local correlation based method for finding the disparity estimation .Experimental results show that our proposed algorithm gives good performance.
Keywords: Stereo Matching, Disparity Map, RMSE Error, Segmentation, k-means.