Specific image feature mining method and system for cervical cancer cell nucleus

The invention provides a specific image feature mining method and system for cervical cancer cell nucleuses. The method comprises the following steps: acquiring a cervical cell sample image; constructing a cell nucleus segmentation model; analyzing and processing a cell nucleus region in the cervical cell sample image by using the cell nucleus segmentation model to extract specific image features; and judging cell types according to the extracted specific image features. The cell nucleus segmentation model is constructed to extract the specific image features of the cell nucleus so as to classify the cells, the classification result is not affected by the size and position of the cells in the image, and the stability and accuracy of the cell classification result are improved; meanwhile, due to the fact that classification is conducted depending on the specific image features, the interpretability and reliability of a cell classification result are high, and the problem that in existing cervical cancer detect