3D Reconstruction Method of Motion Restoration Structure Based on Global and Incremental Estimation
A motion recovery structure and three-dimensional reconstruction technology, applied in computing, 3D modeling, instruments, etc., can solve the problem of low local reconstruction accuracy, achieve low average reconstruction error, improve local accuracy, and improve robustness
Active Publication Date: 2022-05-13
NORTHWESTERN POLYTECHNICAL UNIV
View PDF4 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0004] In order to overcome the shortcomings of the existing 3D reconstruction methods for motion restoration structures, the present invention provides a 3D reconstruction method for motion restoration structures based on global and incremental estimation.
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
[0025] The specific steps of the three-dimensional reconstruction method of motion restoration structure based on global and incremental estimation in the present invention are as follows:
[0026] Step 1. Robust maximum leaf node spanning tree extraction.
[0027] For the input multi-frame images, firstly perform pairwise feature matching, and calculate the relative rotation and relative translation between any two cameras to form the motion information of the two views. Then construct the original scene graph G=(V, E) through the motion information, wherein V is a node set, which represents the camera set of shooting different images in the scene, and the range of the number of nodes contained in it is [300,3000]; E is an edge set, Represents the motion information between pairs of different cameras.
[0028] Extract a robust maximum leaf node spanning tree in the original scene graph G. First, find the point with the largest number of connections from G as a candidate poi...
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 discloses a three-dimensional reconstruction method of a motion recovery structure based on global and incremental estimation, which is used to solve the technical problem of low local reconstruction accuracy in the existing three-dimensional reconstruction method of a motion recovery structure. The technical solution takes the scene graph structure as input, and extracts a robust subgraph as the input of the global estimation, so that a part of high-quality inter-image information can be used to estimate the overall structure of the scene, and the robustness of the estimation can be improved. For the remaining images not added to the scene, the local incremental estimation method is used to add them to the model one by one to improve the local accuracy of model reconstruction. The invention combines the incremental estimation method and the global estimation method, and integrates the advantages of both, so as to achieve a robust and high-precision reconstruction effect. In terms of reconstruction accuracy, the present invention has the lowest average reconstruction error compared with the background art method, and for a data set with 553 images, the average error of the camera position is only 0.3m.
Description
technical field [0001] The invention relates to a method for three-dimensional reconstruction of a motion recovery structure, in particular to a method for three-dimensional reconstruction of a motion recovery structure based on global and incremental estimation. Background technique [0002] As an important topic in the field of computer vision, the 3D reconstruction method of motion restoration structure has great significance in urban modeling, medical applications, virtual reality and other technologies. [0003] Literature "Arie-Nachimson M, Kovalsky S Z, Kemelmacher-Shlizerman I, et al. Global motion estimation from point matches[C] / / 2012 Second International Conference on 3D Imaging, Modeling, Processing, Visualization & Transmission. IEEE, 2012:81-88." A method for estimating 3D structure and camera position using global motion restoration structure is disclosed. It calculates the relative rotation and translation information between two images, first optimizes and ...
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): G06T17/00
CPCG06T17/00
Inventor 张艳宁孙瑾秋李睿朱宇
Owner NORTHWESTERN POLYTECHNICAL UNIV