Display panel microdefect detection method based on two-stage detection network

The invention relates to the technical field of display panel defect detection, and discloses a display panel microdefect detection method based on a two-stage detection network. The detection method comprises the following steps: (1) improving a network structure by using a weighted feature fusion network and a GA-RPN process; (2) expanding the data set by using a Mosaic data enhancement process so as to improve the processing capability of the model on few samples and small defects, inputting the training set after data enhancement into a Faster R-CNN network based on weighted feature fusion for training, and then obtaining model parameters for the panel defect data set; and (3) inputting the defective panel image into the Faster R-CNN network model based on weighted feature fusion, and performing model evaluation and test. Through the weighted feature fusion process, the relation between the deep features and the shallow features is effectively enhanced, the adaptive capacity of the model to different back