Object recognition based on Gabor wavelet and SVM

An object recognition method based on Gabor wavelet and SVM is proposed in this paper. First features of the object are extracted by using Gabor wavelet, and then the dimensions of the Gabor features are reduced with Principal Component Analysis, and finally classification is performed with Support Vector Machine. And this method is applied to the Columbia image library COIL-20 for experiments. Compared to traditional identification methods, experimental results show this method can recognize objects with higher correct recognition rate and less recognition time. That verifies the robustness and effectiveness of the proposed method.