License Plate Detection Based on Fuzzy Rule

Abstract

License plate detection is now-a-days very popular in our country. It also becomes much important to track vehicle as numerous incidents are happening day by day. LP detection is important in places like-parking lot, toll tax collection, trafficmanagement etc.  A fuzzy rule based segmentation technique is proposed in this paper. Before segmentation some pre-processing such as image enhancement, sharpening, noise removal etc. are required to the input image.Input image is segmented by applying fuzzy rules.Rules are based on pixel value of HSV image.It is then binarized andfew post processing tasks are applied like-morphological analysis, region filtering etc. Aspect ratio parameter information helps in extracting the desired region.Ditance to border procedure is tested for verifying shape. Tilt correction is then formed.Finally characters are extracted.  This approach proves remarkable result in LP detection.