Target segmentation method based on non-local feature aggregation neural network

The invention discloses a target segmentation method based on a non-local feature aggregation neural network, and the method comprises the steps: collecting a target video, extracting an original framing image of the video, and obtaining a small sample data set; building a non-local feature aggregation neural network model, and training the non-local feature aggregation neural network model to obtain a target segmentation network; a target video is collected again, and the segmentation quality evaluation score of each target in the image is calculated; judging the image quality according to the segmentation quality evaluation score, continuing to train low-quality images, and retaining target contour key points in high-quality images; and optimizing the key points of the target contour through a BAS-DP lightweight algorithm to obtain a target segmentation result. The method has the advantages of being high in precision, small in training data volume, high in environment adaptability and good in segmentation e