Number Plate Recognition using Image Processing.

Abstract

The main rational of the project is to capture number plate of a moving or a stationary vehicle and extract the number plate information into computer readable ASCII format and send it for subsequent processing depending on the application for verification and authorization. The system based on regular PC environment with a video camera that captures video frames that include car number plates and processes them. The application will be developed on Windows platform using VC++.

Entire effort in the project can be segmented as, Image capturing by a digital camera, Image pre-processing, Edge detection, Image and symbol segmentation and template matching as the final step. Since RMV (Registration Of Motor Vehicles) Sri Lanka has released a new type of number plates which uses a defined format and style, template matching would be a feasible solution in order to recognize elements present in a number plate.

In this scope of work I have assumed that digital images are taken that comprising number plate area and stored in BMP format on 640 X 480 resolution on a storage and mainly concentrated on vehicles which are stationary used in a application such as car parks.