A Novel License Plate Location Method Based On Neural Network And Saturation Information

In a license plate recognition system, the license plate location is a very important step. Itaffects the license plate recognition system's veracity.In this paper, a novel license plate location algorithm for color image is presented. Firstlythe neural networks are used as filters for analyzing within small windows for an image anddeciding whether each window contains a license plate or not. For the template we used beingbig, the regions have many redundancy information. It needs a more accurate method foraccurately locating. We find the information that the license plate's saturation value isdifferent from the backgrounds, so it can be used to locate license plate. But there are stillsome problems that we can't overcome by these two methods. Some effective methods aregiven to resolve these problems. At last, color pairs method is presented to prove that theregion we found is the license plate region or not. The experimental results show thatproposed algorithms are robust in dealing with the license plate location in complexbackground, and the speed is fast.