【摘要】 针对高炉出铁口烟尘浓度动态变化快及传统除尘风机响应滞后等问题，提出一种基于YOLOv8算法的高炉出铁口烟尘图像识别方法。通过集成高分辨率工业相机，构建烟尘动态感知网络，结合YOLOv8s模型实现烟尘浓度分级与扩散形态分割。研究实现了高炉出铁口烟尘的实时精准监测，为钢铁工业污染治理提供智能化解决方案，显著提升环保监管效率并降低安全生产风险。更多还原

【Abstract】 To address the challenges of rapid dynamic changes in dust concentration at the blast furnace taphole and the delayed response of traditional dust removal fans, this study proposes a dust image recognition method based on the YOLOv8 algorithm. By integrating high-resolution industrial cameras, a dynamic dust perception network is constructed, and the YOLOv8s model is employed to achieve dust concentration grading and diffusion pattern segmentation. This research realizes the real-time accurate monitoring of blast furnace taphole dust, provides intelligent solutions for the pollution control of the iron and steel industry, and significantly improves the efficiency of environmental protection supervision and reduces the risk of safe production.更多还原