The invention discloses a
cell image segmentation tracing method based on a U-shaped
residual neural network, and the method comprises the following steps: carrying out the preprocessing of a
cell image, and recording the position of a
cell nucleus according to a cell
data set; generating a moving position vector, and marking a
cell nucleus position and a
time sequence change trend in the cell image; segmenting a cell image and predicting a moving position vector, and simultaneously performing
feature extraction and prediction on the cell image and the moving position vector; post-
processing the moving position vector, reducing over-predicted positions, and interpolating undetected positions; and tracking the segmentation result by the cell, and obtaining a final result by using a
watershed algorithm on the segmentation result and the predicted position. According to the cell
image tracing method, an efficient and easy-to-use segmentation mode is provided, accurate tracing of the cell image on a
time sequence is realized, meanwhile, the
cell segmentation capability is improved, the time of manual labeling of continuous frame cells is shortened, and overall observation in the
cell division and differentiation process in
biology is facilitated.