An Efficient Algorithm for Automatic Car Plate Detection & Recognition

Abstract

From the past few years, vehicle plate recognition is becoming the most challenging and interesting topic of surveillance. It is challenging because number plate comes in different size, shape and written in different font style. It is interesting because of its use in security applications. In this paper we have proposed a robust and simple approach based on edge detection and segmentation with morphological operations. Our suggested approach is performed in three steps: In first step detection of license plate is performed. In second step, segmentation of character is done from detected number plate. While in the third step, recognition of segmented character is performed. We have tested several images taken under different weather condition to check the validity, robustness and efficiency of our algorithm. The result of our experiments is very satisfactory and opens the track of our algorithm to be used in many real life applications.