Image matching method guided by control feature points and optimal local homography in 3D reconstruction of sequential images
A technology of sequence image and 3D reconstruction, applied in image analysis, image data processing, integrated navigator, etc., to achieve the effect of high reliability, high overlap and strong versatility
Inactive Publication Date: 2011-11-30
NAT UNIV OF DEFENSE TECH
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
At present, there is no stable and reliable real-time and fully automatic measurement method for 3D reconstruction of terrain and space targets based on sequential images in my country.
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
[0022] In this embodiment, the image matching method of the control feature points and the optimal local homography guidance in the three-dimensional reconstruction of the sequence image described in the present invention is applied to the three-dimensional reconstruction of the topography and landform by the sequence imaging of the ground by the UAV 。
[0023] The specific implementation steps are:
[0024] The first step is to control feature point selection
[0025] The control feature points of the image in 3D reconstruction need to be evenly distributed. In this step, the first frame of the sequence of images captured by the UAV platform camera on the ground is used as the current key frame, and the control feature points are selected for the current key frame. In order to avoid the distribution of feature points in the direct processing of the image by the interest point extraction algorithm Inhomogeneous problem, using the feature point extraction algorithm of the reg...
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 an image matching method for control feature points and optimal local homography guidance in three-dimensional reconstruction of sequential images, which firstly selects control feature points in the collected sequence images; and tracks the control feature points as the final The feature points of the same name of the key frame image are used to solve the relative geometric relationship between the key frame images; according to the registered control feature points obtained by tracking, the optimal local homography of the point to be matched is calculated, and the optimal local homography is used to predict the to-be-matched The point is located at the position of the feature point with the same name on another sequence image, and is matched and positioned at the predicted position. The invention can be applied to the three-dimensional measurement of the ground by the unmanned aerial vehicle and the restoration of the three-dimensional structure of the space target, and can also be used for the measurement of the scene by the sequence image. At the same time, the present invention enables the matching of feature points with the same name between the key frames of the sequence images to have the advantages of high accuracy, strong anti-interference ability, and fast dense matching speed.
Description
technical field [0001] The present invention relates to technical fields such as camera measurement, three-dimensional reconstruction, space navigation, etc., and furthermore, in the three-dimensional structure reconstruction of topography or space objects based on sequence images, the sequence images are obtained by continuously shooting the ground scene or space objects with a camera, and through sequence The control feature point tracking method of continuous frame images in the image obtains the key frame feature point with the same name and uses the control feature point to calculate the optimal local homography and guide the dense feature point matching method. Background technique [0002] It is an important measurement method in photogrammetry to measure the sequence images formed by continuous shooting of ground scenes and space targets. Obtaining the 3D structure and shape of the target quickly and reliably from the sequence images is the basis for the aircraft to m...
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): G01C11/04G01C21/24G06T7/00
Inventor 李立春刘肖琳于起峰尚洋苑云
Owner NAT UNIV OF DEFENSE TECH