A method for structure recovery from motion based on GPS-assisted information acceleration
A technology of auxiliary information and structural methods, applied in image analysis, image enhancement, instruments, etc., can solve complex and time-consuming problems, and achieve the effect of easy calculation, simple method and fast speed
Active Publication Date: 2021-05-14
QINGDAO RES INST OF BEIHANG UNIV +1
View PDF8 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Some scholars have proposed image matching methods based on graphs and image matching methods based on vocabulary trees. Although they can guide the image matching process, the entire processing process was relatively complicated and time-consuming.
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 present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.
[0026] figure 1 It is a flow chart of the present invention based on the acceleration of GPS auxiliary information to recover the structure method from motion; the factual process of the present invention is mainly divided into the following steps:
[0027]Step 1: numbering the input image and extracting the GPS position information in the input image at the same time,
[0028] (1.1) First calculate a threshold d,
[0029]
[0030] (1.2) Then calculate the image pair to be matched;
[0031] The image i, j satisfying the following conditions is the image pair(i, j) to be matched;
[0032] (wherein, the value of N is taken as 5);
[0033] Step 2: Feature extraction and feature matching of images
[0034] Extract the SIFT features of all the images, and use the cascade hash method to calculate the feature matching for the above-mentioned...
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 recovering structure from motion based on GPS auxiliary information acceleration, including: first step, numbering the input image, extracting the GPS position information in the input image, and calculating the pair of images to be matched ; In the second step, feature extraction is performed on the image, and feature matching is performed between the image pairs to be matched determined in the first step, and the matching results will be used for the 3D point cloud calculation in the fourth step and the 3D point cloud calculation in the third step, respectively. Camera pose estimation; step 3, perform further filtering on the feature matching result of step 2, and then use it to calculate the camera pose; step 4, use the matching result of step 2 and the camera pose calculation in step 3 3D point cloud of the target scene. The present invention utilizes GPS position information to determine image pairs to be matched, and can avoid redundant calculations in the image matching stage to speed up the entire reconstruction process. When the number of images is large, the acceleration effect is more obvious.
Description
technical field [0001] The present invention belongs to the field of image modeling, virtual reality technology, computer vision and computer graphics and image processing, and specifically uses the two-dimensional pixel information of the image of the target scene to recover the three-dimensional image of the scene through theories such as multi-viewpoint geometry and image matching. The structure, including optimizing the shooting position and orientation of the camera, feature point tracking, and the calculation and optimization of the 3D point cloud of the scene, is mainly used for the automatic reconstruction of large outdoor scenes. Background technique [0002] In computer vision, the 3D reconstruction theory of multi-view geometry has been developed relatively maturely, and the relevant basic theories and algorithms are introduced in detail in the document Multiple View Geometry in Computer Vision 2nd. Restoring structure from motion (Structure from Motion, SFM) has ...
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): G06T7/73G06T17/00
CPCG06T17/00G06T2207/10028G06T7/73
Inventor 齐越杜文祥
Owner QINGDAO RES INST OF BEIHANG UNIV