Copper converter blowing end point judgment method based on target detection algorithm

The invention relates to the technical field of copper converter blowing end point judgment, in particular to a copper converter blowing end point judgment method based on a target detection algorithm, and the method comprises the following steps: obtaining a high-temperature melt cooling sample image of copper converter blowing as an input image; the Res2net-ECA module is added into the Yov5s target detection model, and a positioning loss function box loss in an IOU loss function in the Yov5s target detection model is replaced by an Alpha-IOU loss function, so that a Yov5s-Res2net-ECA model is obtained; and the input image is input into the Yov5s-Res2net-ECA model, and the judgment result of the blowing end point of the copper converter is output through the Yov5s-Res2net-ECA model. According to the method, the Res2net-ECA module is added into the Yov5s target detection model, and compared with the original Yov5s model, the number of layers of the neural network is greatly increased, but actually required pa