Research on Traffic Sign Detection Algorithm

Abstract:

As an important part of Advanced Driver Assistance System (ADAS), the traffic sign detection has been paid more and more attention. This paper studied and implemented a valid algorithm of traffic sign detection. Using K-means clustering algorithm to complete the image separation and extraction of prohibition signs from the RGB color image, and then matching them with templates to realize the detection of traffic signs by SIFT algorithm. Series of experiments for traffic sign detection have been carried out to prove the validity and correctness of the algorithm on the basis of the road images in front of the vehicle collected by CCD camera.

• Periodical: Advanced Materials Research (Volumes 945-949)
• Pages: 3304-3308
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.945-949.3304
• Online since: June 2014
• Authors: Mei Hua Xu, Yi Da Liu, Chen Jun Xia
• Keywords: K-Means Clustering Algorithm, Machine Vision, SIFT Algorithm, Traffic Sign Detection

References

[1] Henry X. Liu, Bin Ran: Transportation Research Record (2001), p.161.

[2] Shehroz S. Khan, Amir Ahmad: Pattern Recognition Letters. Vol. 25(2004), p.1293.

[3] Brown M, Lowe D. G: British Machine Vision Conference (2002), p.656.

[4] David G. Lowe: International Journal of Computer Vision. Vol. 60(2004), p.91.

[5] David G. Lowe: International Conference on Computer Vision. Vol. 9(1999), p.1150.

[6] K. Mikolajczyk, C. Schmid: IEEE Trans. Pattern Anal. March. Intell. Vol. 27(2005), p.1615.

[7] Lindeberg, Tony: Applied Statistcs. Vol. 21(1994), p.224.