Performance Analysis of Vehicle License Plate Recognition Using Parallel and Sequential Image Processing on a Multi-Core Processor

Automatic Vehicle License Plate Recognition (VLPR) has become an important application in intelligent transportation systems, traffic management, and security monitoring. However, conventional sequential image processing approaches often experience increased computational time when processing large volumes of visual data. This research presents a performance evaluation of vehicle license plate recognition using sequential and parallel image processing approaches implemented a multi-core embedded processor. The proposed system integrates Python, OpenCV, and enhanced through parallel execution of image processing tasks using the multiprocessing module. The system performs image acquisition, preprocessing, license plate detection, character segmentation, and character recognition using a k-Nearest Neighbour (kNN) classifier. Parallel execution is implemented to accelerate independent processing operations including loading classification datasets and flattened image training data. Experimental results obtained from multiple vehicle license plate images demonstrate that parallel execution significantly reduces execution time compared with sequential processing while maintaining recognition performance. The study confirms that parallel processing improves computational efficiency and supports real-time image processing applications.