Vehicle Number Identification using Machine Learning & OPENCV

Surveillance is important in most of the applications. Vehicle number plate detection is the major part of traffic surveillance. The detection of number plates has become more challenging and interesting in the past few years. The most challenging part of number plate detection is the varying size, shape and font styles of the number plates. The interesting part of number plate detection is its use in security applications. This project proposes a method to detect the vehicle number plate using methods like edge detection and morphological operations. The approach is performed in five steps. The first step is image acquisition which captures the image scene using a camera. Next step is preprocessing which involves conversion of an image to a different model and noise reduction. The next step is license plate detection which involves use of various edge detection algorithms. The final two steps are character recognition and character matching which involves knn classifier and finally the characters are compared with test samples and matched.