A study of the edge detection for road lane

This article presents an investigation of an optimum algorithm for edge detection in order to use in the road lane detection process. The main issues, including the speed, the accuracy, and the limited resources, were taken to consider for the realization on the FPGA technology. The edge detection algorithms of Canny, Prewitt, Sobel and Roberts were compared using MATLAB. A number of road images were captured by a video camera with the image size of 640x480 pixels and the frame rate of 30 fps. In addition, a mask filter was applied to remove red, green, and blue values to help the edge detection process be more efficient. From the experimental results, the Canny algorithm was the most time consuming process, and gave too many lines outside the road lane. Among these, the Roberts algorithm is not only the smallest size, but also gained the fastest speed (3.14 times faster than the Canny algorithm) and the most accurate one to detect the lines of the actual road lanes.

[1]  Rajesh Gupta,et al.  Hardware/software co-design , 1996, Proc. IEEE.

[2]  Liann-Be Chang,et al.  Lane detection system based on software and hardware codesign , 2000, 2009 4th International Conference on Autonomous Robots and Agents.

[3]  Daggu Venkateshwar Rao,et al.  Hardware Acceleration of Edge Detection Algorithm on FPGAs , 2004 .

[4]  B. Michaelis,et al.  Real-time vehicle and lane detection with embedded hardware , 2005, IEEE Proceedings. Intelligent Vehicles Symposium, 2005..

[5]  W. Suntiamorntut,et al.  Hardware/software co-design for line detection algorithm on FPGA , 2009, 2009 6th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology.