Efficient approaches for car license plate localization

Abstract

Automatic license plate recognition alpr is a vital investigation part in the field of intelligent transportation system its it plays a major role in various applications of its. many image processing techniques and systems have been established in recent years to meet the demands for high efficiency alpr system alpr system consists of three major parts such as license plate localization lpl character segmentation cs and character recognition cr lpl is the key part of alpr system it directly involves in the efficiency of the system edge based technique has been one of the major technologies widely used for the detection of license plate lp this thesis explores various difficulties in the detection of lp and proposes suitable methods to overcome the limitations the input images are the car images from india and other countries the major challenges of the lpl are based on non uniform colors non uniform size non uniform characters and non uniform illuminations most of the lpl methods are implemented under constrained input images different features are extracted from the image to find the location of lp edge and color features are simple to construct frequency domain features provide a complex system for the detection of lp the lp area is the high frequency region which is identified by the variation of contrast in the image the high detection rate for various input images captured in different weather conditions is the challenge to alpr system the different constraints such as illumination conditions distance between camera and car clean lp are considered to obtain better detection rate edge based methods are simple and they provide good detection rate for restricted input images newline