Method for assisting in correcting pixel-level noise tag based on RSF model

The invention discloses a method for assisting in correcting a pixel-level noise tag based on an RSF model, and relates to the technical field of medical image segmentation, and the method comprises the steps: inputting an original image and a noise tag; correcting the noise tag based on a plurality of set strategies, and converting the corrected tag into a contour; and carrying out auxiliary selection on the original image and the corrected label based on an RSF model to obtain an optimal correction result, and taking the optimal correction result as the corrected label of the original image. According to the method, two or more strategies for correcting noise labels are set, in the correction process, information of a currently trained original image is used as a selection correction strategy through an RSF model, the optimal corrected label is selected, auxiliary information is provided according to the original image and the contour corresponding to the label, and the correction accuracy is improved. Nois