<![CDATA[This study is motivated by the high rate of traffic violations involving large vehicles such as trucks and buses, particularly on city protocol roads where these types of vehicles are prohibited. Most existing e-ticketing (e-Tilang) systems are still unable to automatically detect large vehicles in real-time and simultaneously read their license plates accurately. This limitation represents a research gap that this study aims to address. The objective of this research is to develop an integrated system that automatically detects violations committed by large vehicles and reads their license plates. The system employs the MobileNet-SSD algorithm for detecting large vehicles from traffic video data and applies the K-Nearest Neighbors (KNN) method for license plate character recognition. Based on tests conducted on five traffic videos under various conditions (morning, afternoon, night), the system achieved a detection accuracy of 90% for large vehicles and an 80% accuracy in license plate recognition. The system performed optimally during daylight hours but showed reduced accuracy at night due to limited lighting conditions. The integration of these two methods has proven to be effective and feasible for real-time law enforcement systems, especially in areas with limited computational resources.]]>
Copyrights © 2025
