FPGA Design and Implementation of Accelerated Stereo Matching for Obstacle Detection

This paper present hardware architecture of low-complexity stereo matching and its implementation. The AD-CENSUS stereo matching algorithm is modified for hardware design and implemented in Xilinx Kintex UltraScale KU040 board. Furthermore the calibration and rectification for stereo camera is included in design. It is accelerated with low hardware resources, and can be obtained the reliable quality of disparity map for obstacle detection.

[1]  Paul A. Viola,et al.  Rapid object detection using a boosted cascade of simple features , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[2]  Yu Wang,et al.  Real-time high-quality stereo vision system in FPGA , 2013, 2013 International Conference on Field-Programmable Technology (FPT).

[3]  Franklin C. Crow,et al.  Summed-area tables for texture mapping , 1984, SIGGRAPH.

[4]  Philip H.S. Torr,et al.  Real-Time Dense Stereo Matching With ELAS on FPGA-Accelerated Embedded Devices , 2018, IEEE Robotics and Automation Letters.

[5]  Zhengyou Zhang,et al.  A Flexible New Technique for Camera Calibration , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Norishige Fukushima,et al.  Fast Implementation of Box Filtering , 2016 .

[7]  A. C. Sonmez,et al.  FPGA design and implementation of a real-time stereo vision system , 2012, 2012 International Symposium on Innovations in Intelligent Systems and Applications.