The spillage on the expressway pavement have the characteristics of centralized distribution, which makes the small target scattered objects in the aerial images of unmanned aerial vehicles overlap, block and lack, which makes it difficult for the traditional target detection algorithm to capture its complex characteristics and can not adapt to the accurate detection of small target spillage. Based on the classic YOLOv8s detection model, an improved detection algorithm is proposed, which conforms to the distribution characteristics of small target spills on expressway pavement. The improved detection algorithm improves the detection accuracy of the algorithm by increasing the detection layer of small target spillage, introducing the cooperative attention mechanism of space and channel, and adjusting the loss function of dynamic non-monotonic focusing mechanism. The empirical analysis results show that the proposed improved detection algorithm has good detection performance, and the detection accuracy is improved by 9.9% compared with the traditional YOLO detection algorithm. Aiming at the sensitivity analysis of attention mechanism and loss function, the improved detection algorithm can improve the detection accuracy and recall rate and reduce the detection time.
Small Target Spillage Detection Algorithm Based on Spatial Attention and Channel Attention Mechanism
Lect. Notes Electrical Eng.
International Conference on SmartRail, Traffic and Transportation Engineering ; 2024 ; Chongqing, China October 25, 2024 - October 27, 2024
Advances and Applications in SmartRail, Traffic, and Transportation Engineering ; Kapitel : 38 ; 374-385
19.07.2025
12 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch