A method to obtain 3D structures from sequential images
A sequence of images, a technology in images, applied in the field of image processing, can solve the problems of cumbersome structure, inability to effectively reflect the 3D characteristics of the target structure, and only reflect the 2D characteristics, so as to achieve the effect of reducing the cumbersomeness
Active Publication Date: 2021-07-20
HUBEI UNIV
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, these methods are based on traditional segmentation algorithms, which not only require manual construction of features, but also mainly reflect 2D characteristics.
The structure is cumbersome and cannot effectively reflect the 3D characteristics of the target structure
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0014] The principles and features of the present invention are described below in conjunction with the accompanying drawings, and the examples given are only used to explain the present invention, and are not intended to limit the scope of the present invention.
[0015] Such as figure 1 As shown, the method for obtaining a 3D structure from sequence images in an embodiment of the present invention includes:
[0016] Step 1, obtain the 2D segmentation result of the target structure in the sequence image according to the deep learning method.
[0017] Step 2, obtaining the 3D connection relationship of the target structure in the z direction according to the 2D segmentation result.
[0018] Step 3: Optimizing the 2D segmentation result of the target structure according to the 3D connection relationship to obtain an optimized 2D segmentation result.
[0019] Step 4: alternately optimize the 3D connection relationship and the optimized 2D segmentation result until convergence,...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention relates to a method for obtaining a 3D structure from a sequence image, comprising: step 1, obtaining the 2D segmentation result of the target structure in the sequence image according to the deep learning method; step 2, obtaining the target structure in the sequence image according to the 2D segmentation result The 3D connection relationship in the z direction; step 3, optimize the 2D segmentation result of the target structure according to the 3D connection relationship, and obtain the optimized 2D segmentation result; step 4, compare the 3D connection relationship and the optimized 2D segmentation result Perform alternate optimization until convergence, and obtain a converged 2D segmentation result and a converged 3D connection relationship; step 5, obtain a 3D visualization result according to the converged 2D segmentation result and the converged 3D connection relationship. The technical solution provided by the invention can effectively acquire the shape and quantity of the target structure from a large amount of sequential images.
Description
technical field [0001] The invention relates to the technical field of image processing, in particular to a method for acquiring 3D structures from sequence images. Background technique [0002] The detection and segmentation of fine structures in sequential images has always been a research hotspot in neurocomputing, which is of great significance to the research of neural circuits, brain neural structure maps, and brain-like intelligence. Therefore, how to obtain the position, shape and quantity of fine structures from micron-scale biological tissues has positive practical significance for humans to understand the connection between their structures and functions. [0003] Since sequence images often have high-resolution characteristics, a small volume can generate a large amount of data, and manual labeling will be a long and boring task. For this reason, research using machine learning and image processing methods to replace manual annotations has emerged. As far as ta...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/00G06T7/11
CPCG06T7/11G06V20/695
Inventor 李伟夫罗志成石畏彭江涛李落清
Owner HUBEI UNIV