Research on the License Plate Recognition based on MATLAB

The license plate image pre-processing which aims to improve the image quality and protrude the outstanding information that we need, which is favourable to the subsequent processing including character segmentation and positioning, character recognition, is significant in the license plate recognition system. First of all, we convert the colour image obtained into a gray-scale image by using histogram equalization, then get rid of the noise through self-adaptive median filter to improve the image quality, and the gray-scale image is stretched to enhance the image contrast. Above all, image is sharpened by using Log-Prewitt operator which does not aim at light intensity directly, but light strong logarithm (light density). to enhance the edge information of the image. It is a new method of Prewitt edge detection with detection boundary area. With light intensity irrelevant, anti-jamming is strong, processing speed is quick and it is suitable for online detection. The experiment results show that the pre-processing method is perfect.