An Efficient Method for Indian Vehicle Number Plate Extraction

Abstract

Vehicle number plate recognition system is the heart of an intelligent traffic system. Extracting the region of a number plate is the key component of the vehicle number plate recognition (VNPR) system. An efficient method is proposed in this paper to analyze using images which often contain vehicles and extract number plate, by finding vertical and horizontal edges from vehicle region. The proposed vehicle number plate detection (VNPD) method consists of 5 stages: (1) Assumption to consider probable region of number plate to concentrate on Region of Interest (ROI), (2) Extracted image from ROI to undergo noise removal and sharpening, (3) Finding vertical and horizontal edges from the image, (4) Applying Connected Component method to extract the number plate, (5) Finally, extracted number plate is subjected for slant correction by Radon transformations. Experimental results show the proposed method is very effective in coping with different conditions like poor illumination, varied distances from the vehicle and varied weather conditions.