Automatic Vehicles Number Plate Identification using Image Segmentation and Comparison Techniques

This paper deals with extraction of textual information from the number plate of vehicles moving on roads based on image processing techniques and retrieval of details of the vehicle from the data base management system. The image of the moving vehicles number plate is grabbed by either using CCD cameras installed on traffic light points or portable/hand held digital cameras with traffic cops. The grabbed image is thresholded to give binary image with white as background and black as textual content. Using segmentation techniques, the text characters are extracted one-by-one. Finally, the characters are compared with the standard shape of the character for its confirmatory test. After, all the characters are identified from the number plate, a text file of the same is generated and fed as input to the data base management system for retrieval of the details of the vehicle under monitor.