Method and system for three-dimensional segmentation of soil pores based on fuzzy clustering

Abstract

An embodiment of the present invention provides a method and system for three-dimensional segmentation of soil pores based on fuzzy clustering. The method includes: reading a group of two-dimensional soil CT images and synthesizing them into a three-dimensional image, and extracting the gray value of voxels in the three-dimensional image Initialize the voxel matrix of the 3D image; preprocess the 3D neighborhood information of the voxel, automatically select the adjacent voxels based on the influence factor of the central voxel based on the 3D neighborhood information; construct the objective function according to the 3D neighborhood information and the influence factor, Fuzzy clustering is performed on the 3D image based on the objective function, and the pore structure identification result is obtained after defuzzification. The method and system for three-dimensional segmentation of soil pores based on fuzzy clustering provided by the embodiments of the present invention, through fuzzy clustering and defuzzification by constructing an objective function based on three-dimensional neighborhood information and influencing factors, has universal applicability to soil CT images and solves the problem of The problem of lack of spatial information and insufficient accuracy in 2D image recognition ensures the execution efficiency of soil pore recognition.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a method and system for three-dimensional segmentation of soil pores based on fuzzy clustering. Background technique [0002] At present, the soil pore identification based on soil CT images mostly uses the FCM algorithm or its improved algorithm FCM_S to identify soil pores based on two-dimensional or three-dimensional images. [0003] When identifying based on two-dimensional images, because the two-dimensional images do not contain spatial information, the three-dimensional neighborhood information is not added for identification, and the connectivity of the three-dimensional structure is lost. The identification results based on two-dimensional images cannot accurately describe the boundaries of complex soil pores. features, leading to inaccurate identification. [0004] When identifying pore structures based on 3D images, the traditional FCM algorithm is used to id...

AI Technical Summary

Problems solved by technology

[0006] In order to solve the problems in existing soil void identification, the embodiment of the present invention provides a method and system for three-dimensional segmentation of soil voids based on fuzzy clustering

Images