Evaluation of dynamic programming among the existing stereo matching algorithms

• Main Authors: Teo, Chee Huat, Nurulfajar, Abd Manap
• Format: Article
• Language: English
• Published: American Institute of Physics (AIP) Publishing LLC 2015
• Subjects: TK Electrical engineering. Electronics Nuclear engineering; QA75 Electronic computers. Computer science
• Online Access: http://eprints.utem.edu.my/id/eprint/14684/1/paper_A00427.pdf; http://eprints.utem.edu.my/id/eprint/14684/; http://scitation.aip.org/content/aip/proceeding/aipcp/10.1063/1.4915817
• Institution: Universiti Teknikal Malaysia Melaka

There are various types of existing stereo matching algorithms on image processing which applied on stereo vision images to get better results of disparity depth map. One of them is the dynamic programming method. On this research is to perform an evaluation on the performance between the dynamic programming with other existing method as comparison. The algorithm used on the dynamic programming is the global optimization which provides better process on stereo images like its accuracy and its computational efficiency compared to other existing stereo matching algorithms. The dynamic programming algorithm used on this research is the current method as its disparity estimates at a particular pixel and all the other pixels unlike the old methods which with scanline based of dynamic programming. There will be details on every existing methods presented on this paper with the comparison between the dynamic programming and the existing methods. This can propose the dynamic programming method to be used on many applications in image processing.