¹School of Information Science and Engineering, Shenyang Ligong University. Shenyang, 110159, P.R. China.
²State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences. Shenyang, 110016, P.R. China.

---

Abstract—According to several key technologies in automatic identification technology of traffic signs, this paper makes a detailed study. First of all, from traffic signs segmentation algorithm, a segmentation algorithm based on iterative segmentation and maximum variance between clusters of traffic signs is studied. Secondly, with feature extraction of traffic signs based on SIFT studied, the codebook is generated by these feature clustering and images are described by histograms using Bag of Words (BOW) model. Finally, multi-class classifier based on SVM is designed to classify traffic signs. The experimental results demonstrate the effectiveness and practicality of the BOW model classification algorithm based on the traffic sign images collected in the natural environment.

Index Terms—Image segmentation, SIFT, SVM classifier, BOW model

[PDF]

Cite: Hongwei Gao, Zhe Liu, Yueqiu Jiang, Bin Li, "Traffic Sign Recognition Technology Based on BOW Model," Journal of Software vol. 8, no. 10, pp. 2652-2659, 2013.