Correlation Based Car Number Plate Extraction System
Car License Plate Recognition (CLPR) is a key research area of computer vision and image processing. And it can be used in various application areas such as security control of restricted areas, traffic law enforcements, surveillance systems, toll collection and parking management systems. In order to recognize the license plate efficiently, location and extraction is the most important part. It determines the accuracy of the whole process. The objective of this work is to detect and extract car number plate area from video clip. So, the input key frames are first generated from the video clip by using key frame generation algorithm. Science this algorithm can reduce the number of frames, the system reduce the processing time. The background region of the plate (Black) and text (White) are usually assigned with opposite colors. By using these connected features of text, this text area is detected and extracted from background with the help of Connected Component Correlation method. The proposed system is tested on 670 samples video clips. These video clips are collected from the internet and public high way lane in Myanmar. The experiments are carried out by using MATLAB programming language. This system can be easily adopted to the system requiring character recognition.
Index Terms- connected component, DWT, horizontal and vertical projection, text extraction, video clip.