Vector field guide fine segmentation method based on coding-decoding network

Abstract

The invention discloses a vector field guided refined segmentation method based on a coding-decoding network, which is characterized in that a convolutional recurrent neural network is adopted to carry out vector field refining and fractional graph prediction on a multi-scale feature graph extracted by a convolutional neural network, and a flow field and a direction field in a vector field are combined to guide refined segmentation. Multi-scale semantic features are utilized, a finer segmentation result is obtained, and the method specifically comprises the steps of image division and gray normalization, network construction and training, image segmentation and the like. Compared with the prior art, the method can fully fuse and utilize the feature information of different scales, can flexibly work on different backbone networks based on CNN, effectively enhances the robustness of different changes in the image, and further improves the segmentation precision.

Description

technical field [0001] The invention relates to the technical field of image segmentation, in particular to a vector field-guided refined segmentation method based on an encoding-decoding network. Background technique [0002] Image segmentation technology is one of the foundations of computer vision, and it is also one of the difficulties in semantic understanding of images. With the vigorous development of deep learning theory and the continuous growth of computing resources, the efficiency and accuracy of image segmentation have been greatly improved. Long et al. proposed the fully convolutional neural network (FCN) in 2015, which modified the last fully connected layer of the general classification network into a convolutional layer, and adopted a point-by-point addition strategy in the process of feature fusion; in the same year, Navab et al. proposed U-Net that stitches and fuses features at the channel level; after that, He Yuming et al. proposed the residual represe...

AI Technical Summary

Problems solved by technology

However, the current existing work needs to improve the fusion method of different scale features, and there is still information loss in the processing process, and because multi-scale features can be regarded as sequence data, RNN is more capable of processing in a hierarchical structure than CNN Advantages of multi-scale features

Images