A three-dimensional image segmentation method and system based on a full convolutional neural network

Abstract

The invention discloses a three-dimensional image segmentation method and system based on a full convolutional neural network, and the method comprises the following steps: step1, collecting and obtaining a sequence image, and carrying out the marking of the sequence image, and obtaining training sample data; step 2, performing normalization preprocessing on the training sample data obtained in the step 1; step 3, applying the sample data processed in the step 2 to carry out supervised training on pre-constructed 3-D full convolution residual U-net network model, training to a preset convergence condition, and obtaining a trained three-dimensional image segmentation model; and step 4, after normalization processing is carried out on the sequence image data to be segmented, inputting the sequence image data into the three-dimensional image segmentation model trained in the step 3, and obtaining a sequence image segmentation result. The continuity information of the sequence can be fullyutilized, and a relatively good result can be obtained in three-dimensional image segmentation.

Description

technical field [0001] The invention belongs to the technical field of three-dimensional sequence image segmentation, in particular to a three-dimensional image segmentation method and system based on a fully convolutional neural network. Background technique [0002] In the field of image processing, it generally includes image segmentation, registration, fusion and 3D reconstruction processes. In the field of image segmentation, the existing technology mainly analyzes each layer of images in a sequence, and uses traditional graphics methods, machine learning methods and deep learning methods to segment on a two-dimensional plane, and then perform three-dimensional image segmentation. Fusion to achieve 3D image segmentation. The image sequence segmentation performed by the existing methods generally only considers the relationship between the pixels on the two-dimensional level, while ignoring the continuity between the images of each layer, thus losing more context inform...

AI Technical Summary

Problems solved by technology

Segmentation of image sequences by existing methods generally only considers the relationship between pixels on the two-dimensional level, while ignoring the continuity between images in each layer, thus losing more context information

[0003] For example, the traditional algorithm based on threshold segmentation is simple in principle, and image segmentation is achieved by manually traversing to select the best threshold; however, its calculation process is complex, susceptible to noise interference, and poor in robustness

The algorithm based on edge detection is to detect the edge points in the picture first, and then connect them into contours according to a certain strategy to form a segmented area; its disadvantage lies in the contradiction between noise resistance and detection accuracy, so the obtained segmentation is often intermittent and inconsistent. complete structural information

Images