<![CDATA[The rapid increase in vehicle numbers has rendered traditional manual traffic monitoring approaches inefficient and unreliable, thereby emphasizing the need for intelligent, automated systems to assist in traffic management and law enforcement. Among these, Automatic License Plate Recognition (ALPR) systems play a crucial role in detecting and tracking vehicles. However, their performance often deteriorates under low-light or poor visibility conditions, leading to reduced detection accuracy. To address this limitation, this study proposes a two-stage image enhancement pipeline that integrates Contrast Limited Adaptive Histogram Equalization (CLAHE) and Unsharp Masking (USM) techniques with the advanced YOLOv11 object detection model. The dataset utilized comprises 1,496 images extracted from Electronic Traffic Law Enforcement (ETLE) video footage captured in Makassar, Indonesia. These images were systematically divided into training, validation, and testing sets in a 70:20:10 ratio to ensure balanced model evaluation. Four experimental scenarios were conducted to assess the contribution of each enhancement method. The results revealed that the combined application of CLAHE and USM significantly improved detection accuracy, achieving a Precision of 0.945, Recall of 0.977, and a mean Average Precision (mAP@0.5:0.95) of 0.830—outperforming all other configurations. These findings confirm that the synergistic use of contrast enhancement and edge sharpening effectively mitigates the challenges posed by low-light environments. Consequently, the proposed approach enhances the robustness, clarity, and reliability of ALPR systems, offering a practical solution for real-world intelligent transportation applications and automated traffic law enforcement.]]>
