An Efficient Relative Gradient Based Radiometric Invariant Stereomatching Using Guided Filter
Sarika S1, Deepambika V. A2, M. Abdul Rahman3
1Sarika S, Department of Electronics and Communication, LBS Institute of Science and Technology for Women, Trivandrum (Kerala), India.
2Deepambika V. A, Department of Electronics and Communication, LBS Institute of Science and Technology for Women, Trivandrum (Kerala), India.
3Dr. M. Abdul Rahman, Professor, Vice Chancellor of Kerala Technical University, Trivandrum (Kerala), India.
Manuscript received on 15 August 2015 | Revised Manuscript received on 25 August 2015 | Manuscript Published on 30 August 2015 | PP: 105-109 | Volume-4 Issue-6, August 2015 | Retrieval Number: F4180084615/15©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© 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: Stereomatching algorithms provide a better disparity map only when the stereo image pairs under consideration are under similar radiometric conditions, but under real world scenarios this condition may not hold. As a result of this corresponding pixels in the left and right image will be at different intensities and most of the state of the art stereo matching algorithm fails to provide a better disparity map. To overcome this issue this paper proposes a relative gradient based approach. Also in order to have better edge preservation and faster result guided filter based cost aggregation is used. The result shows that the proposed method performs well under varying radiometric conditions where the conventional state of the art stereo matching algorithms fail.
Keywords: Stereomatching, Radiometric Variations, Relative Gradient, Guided Filter
Scope of the Article: Radio Over Fiber