Edge Detection Algorithm Based on YOLOv8

With the rapid development of computer vision technology, edge segmentation has garnered significant attention as a key technique in image processing. Traditional edge detection algorithms often struggle to handle complex scenarios and diverse target morphologies, while deep learning models have provided new solutions. In this study, we proposed an edge segmentation method based on the YOLOv8 model, which achieves efficient identification of target edges in the complex scenario of rotary kilns through improved network architectures and training strategies. Experimental results demonstrate that the method performs exceptionally well across multiple evaluation metrics, showcasing high practical value.