Exploring Gender Dynamics in Motorcycle Traffic Violations Through Gender Detection Technology

Urban traffic management faces significant challenges due to non-compliance with traffic rules, particularly among motorcycle riders. This study introduces an innovative approach employing the YOLOv9 object detection framework to monitor and analyze motorcycle violations at busy intersections in Marrakech. By leveraging advanced machine learning techniques, the study aimed to detect helmet usage, traffic light violations, and assess gender-based differences in violation rates. Data were collected using smartphone cameras at strategic locations during peak traffic times across selected days. The captured video footage was analyzed using the YOLOv9 model, which was pretrained on diverse traffic scenes to enhance its accuracy and reliability in real-time object detection. The findings reveal that the automated system not only aligns closely with manual counting but also offers greater consistency and efficiency. Key results indicated a pronounced gender disparity in violation frequencies and provided insights into the temporal patterns of traffic rule infractions. Moreover, the study highlighted minor discrepancies due to environmental factors, which were systematically addressed to refine the detection process. Contextual factors in interpreting detection results.