Share
Export Citation
Improvement of Vehicle License Plate Detection and Recognition with the Addition of the CLAHE Method
Zulkarnain E.
Beyond Technology Summit on Informatics International Conference Bts I2c 2025
Abstract
Automatic License Plate Recognition (ALPR) detection and recognition play an important role in modern traffic monitoring systems, especially in the implementation of Electronic Traffic Law Enforcement (ETLE). The main challenge of this system is the low image quality owing to uneven lighting and vehicle movement. This study proposes to improve the performance of the YOLOv11 model by adding a preprocessing stage using Contrast Limited Adaptive Histogram Equalization (CLAHE) and Unsharp Masking to improve contrast and image sharpness. The dataset was taken from ETLE CCTV footage on Jalan A.P. Pettarani, Makassar, under low-light conditions. The test results showed that the combination of CLAHE and Unsharp Masking gave the best performance with a Precision of 0.946, Recall of 0.977, mAP50 of 0.978, and mAP50–95 of 0.833. The evaluation of character recognition using EasyOCR also showed the lowest Character Error Rate (CER) value of 6.35%. Thus, this combined method has been shown to be effective in improving the detection and recognition accuracy of license plates and has the potential to be implemented in ETLE systems under real traffic conditions.