Key points are not available for this paper at this time.
Objectives: This study presents a resilient deep learning method for vehicle detection and classification in traffic aerial images. Methods: The Roboflow dataset used for model training includes 220 aerial images, annotated in YOLOv8 format in four classes of moving objects. Our suggested YOLO version eight-based approach uses real-time photos from Aerial cameras for the training and testing and validation for the classification and detection of moving objects. Findings: YOLOv8, a state-of-the-art object detection algorithm, enhances the accuracy and efficiency of traffic monitoring on the Roboflow dataset. This approach achieves a mean average precision (mAP) of 0.86, demonstrating the robustness of our model in detecting traffic-related objects. The precision and recall values of 0.89 and 0.96, respectively, further affirm the model's effectiveness in accurately identifying and classifying objects. Novelty: The research introduces a novel application of YOLOv8 for traffic analysis, demonstrating its advanced object detection and classification capabilities and achieving an impressive mean average precision of 0.86, enhancing computer vision for real-time traffic monitoring, which offers high precision and recall, enabling traffic flow optimization, anomaly detection, and intelligent transportation systems, paving the way for smart city solutions and data-driven urban planning. Keywords: Object Detection, YOLOv8, RealTime Monitoring, Intelligent Transportation Systems, Aerial Images
Pandey et al. (Sat,) studied this question.