A New Vehicle Recognition Approach Based on Graph Spectral Theory and BP Neural Network

A new vehicle recognition approach based on graph spectral theory and neural networks is proposed in this paper. In the approach, image threshold method based on graph spectral theory is used for image preprocessing. And after filter of undetermined regions with rules, regions left are gray-unified. These gray values are input into neural network to recognize vehicle and vehicle types. The experiment proves that this method has high recognition rate and low false rate.