Traffic sign recognition using dual tree-complex wavelet transform and 2D independent component analysis

A novel traffic sign recognition algorithm is presented in the paper. This algorithm integrates the Dual-Tree Complex Wavelet Transform(DT-CWT) representation of traffic sign images and 2D Independent Component Analysis(2DICA) method. First traffic sign color-image is preprocessed with gray scaling, and normalized to 64×64 size. Image features could be obtained by concatenation of four levels DT-CWT images which are used to represent gray image of traffic sign. Second, 2DICA and a nearest neighbor classifier are used to recognize the traffic signs. The whole recognition algorithm is implemented for classification of 50 categories of traffic signs, and its accuracy reaches to 97%. It also compares the presented algorithm with well-established image representation like template, Gabor, and feature selection techniques such as PCA, LPP, 2DPCA at same time. Experimental results indicate that the proposed algorithm was robust, effective, and accurate to recognize traffic signs.