The Implementation of Tesseract OCR Engine for Automated Plate Number Coding Violation Detection Tool

Abstract

The main objective of the study is to develop an automated plate number coding violation detection tool using Tessaract Algorithm. The study utilized the descriptive and experimental method of research. In this research, the proponent investigated the current processes on the Non-Contact Apprehension Program of the Metro Manila Development Authority (MMDA). Interviews were done with the head and personnel of the No Contact Group and the production data as well as the actual video coverage for pilot cameras/CCTV were also obtained. As to the experimental method of research, random images were captured from the video samples and then inputted to the proposed system to determine accuracy and processing time to convert the captured plate number images into text forms where it can be collected and saved on a database. Findings revealed that the proposed system or tool could detect the plate numbers of most of the vehicles especially those of the new series plates whose fonts are in black and with a white background while those plates with a background image and light green-white background and using green fonts have a high failure rate of detection and text conversion. Plates of motorbikes and induction stickers were not detected as well as those vehicles with plate covers or customized plates.