Estimation of disparity map of stereo image pairs using spatial domain local Gabor wavelet. Issue 4 (1st August 2015)

• Record Type: Journal Article
• Authors: Malathi, T.; Bhuyan, M.K.
• Is Part Of: IET computer vision. Volume 9:Issue 4(2015)
• Page Start: 595
• Page End: 602
• Publication Date: 2015-08-01
• Subjects: stereo image processing -- Gabor filters -- image matching -- cameras -- image capture -- image filtering -- median filters; stereo image pair disparity map estimation -- spatial domain local Gabor wavelet -- image capturing -- cameras -- quantitative disparity information recovery -- image occlusion -- image discontinuity -- texture-less image -- feature-based stereo matching method -- feature-based stereo correspondence -- two-pass cost aggregation step -- disparity computation step -- winner-takes-all selection -- disparity refinement step -- Kuwahara filter -- median filter -- middlebury benchmark database; Computer vision -- Periodicals; Pattern recognition systems -- Periodicals; 006.37
• DOI: 10.1049/iet-cvi.2014.0210
• Languages: English
• ISSNs: 1751-9632

Abstract

Abstract : The stereo matching problem takes two images captured by nearby cameras and attempts to recover quantitative disparity information. Most of the existing stereo matching algorithms find it difficult to estimate disparity in the occlusion, discontinuities and textureless regions in the images. In the last few decades, a number of stereo matching methods have been proposed to overcome some of these problems. In the same line of thought, the authors propose a new feature‐based stereo matching method, which consists of four basic steps – feature‐based stereo correspondence, two‐pass cost aggregation, disparity computation using winner‐takes‐all selection and finally, the disparity refinement. In the proposed method, local features of Gabor wavelet in spatial domain are used for matching cost computation and subsequently a cost aggregation step is implemented by combined use of the Kuwahara filter and the median filter. Experimental results on the Middlebury benchmark database shows that the proposed method outperforms many existing local stereo matching methods.