Three-dimensional image segmentation method and system, storage medium and electronic equipment

Abstract

The invention discloses a three-dimensional image segmentation method and system, a storage medium and electronic equipment, and the method comprises a supervoxel result obtaining step: carrying out initial segmentation of a three-dimensional image through a supervoxel generation algorithm, and obtaining supervoxels; a super voxel feature matrix obtaining step: calculating the features of the super voxels by taking the super voxels as units, and then performing aggregation to obtain a super voxel feature matrix; a clustering result obtaining step: obtaining a clustering result through a self-supervised deep subspace clustering network model according to the super voxel feature matrix; and an image segmentation step: after mapping the clustering result back to the three-dimensional image, completing image segmentation of the three-dimensional image. Therefore, on the basis of the super voxel, automatic segmentation of the three-dimensional image is completed under the condition of not using annotation data.

Description

technical field [0001] The present invention relates to the field of three-dimensional image segmentation, in particular, to a method, system, storage medium and electronic equipment for three-dimensional image segmentation based on self-supervised depth subspace clustering. Background technique [0002] Image segmentation (imagesegmentation) technology is an important research direction in the field of computer vision and an important part of image semantic understanding. Image segmentation refers to the process of dividing an image into several regions with similar properties. From a mathematical point of view, image segmentation is the process of dividing an image into mutually disjoint regions. In recent years, with the gradual deepening of deep learning technology, image segmentation technology has developed by leaps and bounds. Technologies related to this technology, such as scene object segmentation, human body front and background segmentation, human face parsing, a...

AI Technical Summary

Problems solved by technology

[0012] 1. The segmentation method based on graph theory is very dependent on the contrast of color grayscale information between pixels or voxels in the image. When encountering some blurred images with small color grayscale differences, the processing results are not ideal.

[0013] 2. The clustering-based method is greatly affected by the initial segmentation results. The rough clustering results for the first time may be different each time, making the results of subsequent iterations different, so this type of method is not stable enough

However, for image processing tasks in real life, the cost of a large amount of labeled data is high, which limits its use scenarios

Images