A Vehicle License Plate Location Algorithm Based on Image Edge Features and Neural Network

Automatic location of vehicle license plates is a difficult image segmentation problem.Neural network is a powerful tool for solving the problem.In this paper,an algorithm based on neural network and image edge features is presented.It gets out the samples through a slide window and puts them into the neural network.The output eigenvectors of a BP net is used to describe the characteristics of edges at various places where can be marked with the top left corner of the slide window in an image.Combined with statistics,this algorithm can detect possible license plates in an image very precisely and efficiently.