Design of Hough transform hardware accelerator for lane detection

Abstract

The Hough transform is a well-known straight line detection algorithm and it has been widely used for many lane detection algorithms. However, its real-time operation is not guaranteed due to its high computational complexity. In this paper, we designed a Hough transform hardware accelerator on FPGA to process it in real time. Its FPGA logic area usage was reduced by limiting the angles of the lines to (-20, 20) degrees which are enough for lane detection applications, and its arithmetic computations were performed in parallel to speed up the processing time. As a result of FPGA synthesis using Xilinx Vertex-5 XC5VLX330 device, it occupies 4,521 slices and 25.6Kbyte block memory giving performance of 10,000fps in VGA images(5000 edge points). The proposed hardware on FPGA (0.1ms) is 450 times faster than the software implementation on ARM Cortex-A9 1.4GHz (45ms). Our Hough transform hardware was verified by applying it to the newly developed LDWS (lane departure warning system).