Developing a preprocessing detection and recognition method for transport and communication sector: The English Iraqi car plates as a current study

Intelligent transportation systems (ITSs) applications have an important role in extending security to the transportation sector. The most important example of these applications is the automatic number plate recognition system. It helps to monitor the traffic road and apply protection and safety to enforce the law. Although several different methods have been proposed to detect car plate numbers from car images or a series of car images in the video clip, plate detection techniques still need an efficient method with good accuracy and low runtime for real-time detection systems. For the detection phase, new preprocessing methods have been used, such as binarization, filtering, dilation, erosion, convolution, flood fill, image cropping, and smearing. After that, the template matching method has been used for the car number plate recognition phase. This method evaluates the car plate image pattern that has been recognized and compares it with the car plate image pattern in the template stored in the database. Based on the tests that have been done on 100 English Iraqi car images, the detection accuracy is 94% with a recognition accuracy of 91%. The execution time of the entire detection and recognition system is 2ms per car image. The results obtained can be used for implementation in any security system, for example, car parking management systems.