Car License Plate Detection and Recognition using Morphological Image Processing and Template Matching

Abstract

This study presents a robust method for vehicle number plate detection and identification. It mostly consists of three stages: plate extraction, plate number separation, and number recognition. Morphological operation is used to extract plate and numbers, whereas template matching is used to recognize the extracted number images. After that, it looks into a predefined dataset to decide if the number is licensed. Being independent of plate size, color and illumination make the system robust explicitly. Depending on camera position, it can be executed in two modes: mode 1 for near images while mode 2 for rear images. Though mode 2 presents better performance for rear images, it is somewhat slower in operation compared to mode 1. Experimental result also shows the effectiveness of the system even for images taken from an angular position. So, it can easily be implemented at car parking center, toll collecting point or at light post to capture images.

Keywords: Image processing, license plate, morphological operation, template matching, MATLAB Cite this Article Md. Ashraful Islam, Syed Iqram Hussain, Rafiqul Islam.

Car License Plate Detection and Recognition using Morphological Image Processing and Template Matching. Journal of Computer Technology & Applications. 2020; 11(1): 21–27p.