Characters feature based Indian vehicle license plate detection and recognition

Vehicle license plate recognition is well known system comes under machine vision which employed mostly for traffic management, preventing vehicle crime, parking system, road tolling which can able to recognize number from vehicle license plate. This paper present an approach which is based on characters of the number plate, for that an adaptive preprocessing method was used because not every time the illumination condition is constant, then segmentation of character were done using vertical and horizontal projection of the extracted license plate. Finally character recognition were done using K-nearest neighbor classifier based on feature vector which was extracted from the boundary analysis of the character. This system was tested on various vehicle images taken in different illumination condition.