Improved YOLOv8-based Vehicle Detection Method for Road Monitoring and Surveillance

To address the problem that it is difficult to balance the accuracy and speed of vehicle detection in current traffic monitoring scenarios, this paper proposes a lightweight vehicle detection algorithm based on YOLOv8. First, the detection speed and accuracy of the model are improved by replacing the Neck layer of the original model with slimneck; second, the SimAM attention mechanism is introduced to strengthen the key features of the vehicle and suppress the non-key features; and finally, the WIoU loss function is adopted to achieve lower regression error. The experimental results on the UA-DETRAC dataset show that, compared with the base model, the improved method in this paper improves 2.66% and 7.14% in the two indexes of mAP and FPS, which effectively improves the problem of lower vehicle detection accuracy in complex traffic scenarios and achieves faster detection speed.