Traffic sign recognition using HOG-SVM and grid search

Abstract

Considering the lower accuracy of existing traffic sign recognition methods, a new traffic sign recognition method using histogram of oriented gradient - support vector machine (HOG-SVM) and grid search (GS) is proposed. First, the histogram of oriented gradient (HOG) is used to extract the characteristics of traffic sign. Then the grid search technique is applied to optimize the parameters of support vector machine (SVM). Finally, the traffic sign is recognized by using the trained SVM classifier. Experimental results indicate that the proposed method could achieve high accuracy for traffic sign recognition.