A NOVEL APPROACH FOR AUTOMATIC NUMBER PLATE RECOGNITION SYSTEM FOR INDIAN NUMBER PLATES

This paper presents the Automatic Number Plate Recognition (ANPR) System for Indian Cars. ANPR is a real life application. Basically ANPR systems are consist of Vehicle image capture, Preprocessing, Extraction of number plate, Character segmentation and Character recognition steps. Image capturing is usually done by existing closed-circuit television or road-rule enforcement cameras, or special cameras specifically designed for the task. ANPR systems should be quickly and should process license plates under different environmental conditions, such as indoors, outdoors, day or night time successfully. We are using me dian filter, histogram equalization to pre-process the vehicle image. Segmentation technique is based on discontinue properties of pixels in which we detect points, lines and edges of number plate. For edge detection we use “Canny Edge Detector” operator. And for character recognition we have applied SVM technique. Experiments are performed using database of car images and the method gives 98.33% correct classification rate.