Enhanced YOLOv8 for Forest Fire Detection: Integrating SlimNeck-MCA Architecture and MPDIoU Loss Function

In order to address the problems such as frequent occurrence of forest fires, insufficient promptness of alarm, and difficulty in detection, this paper proposes an optimized algorithm based on YOLOv8. By introducing SlimNeck lightweight bottleneck structure, MCA multi-dimensional attention mechanism and MPDloU loss function, the detection accuracy and real-time performance of the algorithm have been significantly improved. Experiments show that the improved model achieves a mean average precision Map of 75.4% at a resolution of 640×640, which is 1.6% higher than that of the original model. The parameter size reaches 2.66M, which is 11.6% lower than that of the original model, and the inference speed has also been greatly enhanced. Finally, through ablation experiments, the effectiveness of each improvement module was verified, providing a better solution for forest fire early warning.