Review of stereo matching algorithms for 3D vision
2007 (English)In: 16th International Symposium on Measurement and Control in Robotics, 2007Conference paper (Refereed)
Stereo vision, resulting in the knowledge of deep information in a scene, is of great importance in the field of machine vision, robotics and image analysis. As a result, in order to address the problem of matching points between two images of a stereo pair several algorithms have been proposed so far. In this paper, an explicit analysis of the existing stereo matching methods, up to date, is presented in full detail. The algorithms found in literature can be grouped into those producing sparse output and those giving a dense result, while the later can be classified as local (area-based) and global (energy- based). The presented algorithms are discussed in terms of speed, accuracy, coverage, time consumption and disparity range. Comparative test results concerning different image sizes as well as different stereo data sets are presented. Furthermore, the usage of advanced computational intelligence techniques such as neural networks and cellular automata in the development and application of such algorithms is also considered. However, due to the fact that the resulting depth calculation is a computationally demanding procedure, most of the presented algorithms perform poorly in real-time applications. Towards this direction, the development of real-time stereo matching algorithms, able to be efficiently implemented in dedicated hardware is of great interest in the contexts of 3D reconstruction, simultaneous localization and mapping (SLAM), virtual reality, robot navigation and control. Some possible implementations of stereo- matching algorithms in hardware for real-time applications are also discussed in details.
Place, publisher, year, edition, pages
Computer Vision and Robotics (Autonomous Systems)
IdentifiersURN: urn:nbn:se:kth:diva-51028OAI: oai:DiVA.org:kth-51028DiVA: diva2:463256
16th International Symposium on Measurement and Control in Robotics 21-23 June 2007 - Warsaw, POLAND
QC 201112092011-12-082011-12-082011-12-09Bibliographically approved