Automatic Vehicle License Plate Segmentation And Recognition Based On Color Detection

Vehicles are used in our daily lives for transporting people and cargo.Automatic vehicle license plate segmentation and recognition(AVLPSR)is one of the key applications of the intelligent transportation system(ITS).Now,AVLPSR sy'stems are used in various real-life applications such as parking lot,automatic toll collection,access control.traffic law enforcement,border crossing control,and traffic monitoring,etc.AVLPSR systems are working properly and give good performance in almost every field.There are many types of standard vehicle license plates in China which include 38 Chinese characters,24 English alphabets(except O&I)and 10 numbers(except for provisional vehicle license plates).AVLPSR method uses an image processing technique to detect license plates from vehicle images.The most important and difficult step in the AVLPSR method is segmentation of a vehicle license plate that straightly disturbs the entire proficiency of the method.The manifestation of blur and noise in an image,dirt,rain,improper illumination,fog,and less bright condition,etc.make the job more difficult.The license plate segmentation method is used for identifying the vehicle license plate from the image and induce important data from the license plate.Automatically recognize the license plate from a vehicle license plate image is dif'ficult because of challenges concerned in the application domain.The main challenges observed in the license plate are color,font,size,occlusion,location,and multi-plats.This AVLPSR proposed research work is mainly focused on detection,segmentation,and recognition of Chinese vehicles license plate.The main process of AVLPSR is divided into three steps:vehicle license plate detection,character segmentation,and character recognition.In this work a simple algorithm,local binary pattern histogram(LBPH)is used for vehicle license plate detection and we will get ideal results through this method.For segmentation,connected components(CC)algorithm is proposed.The proposed algorithms have two advantages.First,if the Chinese character of a vehicle license plate consists of only one connected stroke/part,this algorithm segments the vehicle license plate into required eight characters as Chinese standard license vehicle plate has eight characters(mostly one Chinese character,two/three English alphabets,three/four digits and a dot at third place).Secondly,if a Chinese character has more than one stroke/part then this algorithm merges those strokes iinto one character.This proposed algorithm gives satisfactory results for all alphanumeric characters and especially for Chinese characters.For recognition of a license plate,support vector machine(SVM)is used.The results of the proposed work demonstrate that license plate detection and segmentation methods perform better in terms of accuracy and performance and achieved an accuracy rate of recognition is 96.7%.