Vehicle registration and notification system for CCTV on pavement

Abstract

The research presents an image processing technique to identify the license plate of cars/motorcycles that drive on sidewalks. The model adapted footage that recorded from as a sample data in the development of analytical models of 3 important cases: 1) Detection of vehicles driving on the sidewalk 2) Detection of license plate locations 3. Comparison of numbers Information on the vehicle registration and compiled into a sentence

Initially, the researcher used the image to create a pedestrian template with a car on the sidewalk and a road without a car on the sidewalk to let the program learn what the user wants to detect. When detected, the program was designed to use the border inspection technique to specify the position of the license plate. After that, the detected images will be processed into the 3rd digital image processing process to compare and compile into the final characters of this research. The results show that the data analysis model can detect motorcycles and cars driving on the sidewalk. On the case that there is no precise person walking on the road Can detect the license plate at the back of the car and can compare images with characters and compile them into characters but the accuracy of the character comparison depends on the quality of the image including light-shadow, the integrity of the license plate, which affects the accuracy in comparing characters.