Low-Cost Stereo Vision on an FPGA

We present a low-cost stereo vision implementation suitable for use in autonomous vehicle applications and designed with agricultural applications in mind. This implementation utilizes the Census transform algorithm to calculate depth maps from a stereo pair of automotive-grade CMOS cameras. The final prototype utilizes commodity hardware, including a Xilinx Spartan-3 FPGA, to process 320times240 pixel images at greater than 150 frames per second and deliver them via a USB 2.0 interface.

[1]  R. Zabih,et al.  Individuating unknown objects by combining motion and stereo , 1994 .

[2]  Ramin Zabih,et al.  Non-parametric Local Transforms for Computing Visual Correspondence , 1994, ECCV.

[3]  Richard Szeliski,et al.  A Taxonomy and Evaluation of Dense Two-Frame Stereo Correspondence Algorithms , 2001, International Journal of Computer Vision.

[4]  John Woodfill,et al.  Real-time stereo vision on the PARTS reconfigurable computer , 1997, Proceedings. The 5th Annual IEEE Symposium on Field-Programmable Custom Computing Machines Cat. No.97TB100186).

[5]  Reinhard Männer,et al.  Calculating Dense Disparity Maps from Color Stereo Images, an Efficient Implementation , 2004, International Journal of Computer Vision.

[6]  Darius Burschka,et al.  Advances in Computational Stereo , 2003, IEEE Trans. Pattern Anal. Mach. Intell..