Metal corrosion degree calculation method based on pixel clustering Transform model algorithm

The invention discloses a metal corrosion degree calculation method based on a pixel clustering Transform model algorithm, and the method comprises the steps: generating a pseudo tag through a k-means clustering algorithm, carrying out the adaptive training through a pixel clustering Transform model, inputting a to-be-detected metal cross section image into the trained model, and finally obtaining the recognition result of a metal corrosion part and the metal corrosion degree. And calculating to obtain the metal corrosion degree. The method is used for detecting the metal corrosion part and calculating the metal corrosion degree, and compared with the prior art, the method can realize full-automatic processing after image loading; the efficiency of daily diagnosis of metal corrosion conditions by related industrial personnel can be improved, and errors are greatly reduced. 本发明公开了一种基于像素聚类Transformer模型算法的金属腐蚀度计算方法，利用k-means聚类算法生成伪标签，然后通过像素聚类Transformer模型进行自适应训练，将待测金属截面图像输入到训练好的模型中，最终得到金属腐蚀部分和金属腐蚀程度的识别结果，计算得到金属腐