The use of mathematical morphology and template matching in automatic malaysian car plate number recognition

Car plate number nowadays is an essential component to be recognized because of the useful of its widely application especially in this technological era. Therefore, a method of detection, extraction and recognition of car plate image are important in order to be applied in various applications. Presently, the existing methods have several problems that lead to this study which are high time consuming for training purposes and requirement of complex steps in preparing the whole car plate recognition process. Therefore, the following objectives are set up that are to implement recognition method using Mathematical Morphology and Template Matching on car image which contain plate number. Next, is to evaluate the accuracy of the method. In implementing this study, 25 images data of car with its plate number are collected and used. Mathematical Morphology is used to detect the edge of each character on the car plate while the Template Matching as a tool to match up the detected and template character in order to recognize the registered car plate number. The accuracy of the method is evaluated using percentage recognition rate. Results show that the performance of the method successfully extracted the car plate number area by 92% of the original image. Then for the other of 23 extracted images, it was 86.95% that successfully recognized the number plate. All of these results create its own novelties which are the ability to reduce processing time and having high detection and recognition rate. In conclusion, the proposed method can be one of the alternatives in car plate recognition process in the meanwhile it can be used widely in digital data transfer such as Automatic Enforcement System (AES), automatic saved car park number in smart parking system and high technological verification application of home security system.