Method for improving three-dimensional reconstruction point-clout density on the basis of contour validity

Abstract

The invention discloses a method for improving three-dimensional reconstruction point-clout density on the basis of contour validity. The method comprises the following steps: 1, extracting an object contour, and generating a corresponding effective area graph sequence; 2, calculating expansion dimensions of a point cloud on an x axis, a y axis and a z axis; 3, obtaining a derivative point cloud by expanding each point in an initial point cloud; 4, converting the derivative point cloud to a camera coordinate system, performing back projection on the derivative point cloud to an effective area graph, and reserving points in an effective area; 5, calculating point products of vectors from initial points of processed derivative points to the points and normal vectors of the points, and reserving points whose point product values are greater than zero; and 6, inspecting whether density of the derivate point cloud reaches a demand, and if the density does not satisfy the demand, by taking the derivative point cloud as the initial point cloud, carrying out operation after the second step until the demand is satisfied. According to the invention, the method is not restricted to a specific detour shot image sequence, does not excessively rely on parameter adjustment, and can improve the density of an effective point cloud within quite short time under the cognition of a quite small computation amount.

Description

technical field [0001] The invention relates to the field of computer vision, in particular to a method for improving the density of a three-dimensional reconstruction point cloud based on contour effectiveness, which can effectively increase the density of a point cloud in a relatively fast time. Background technique [0002] Computer vision involves many disciplines and is the inverse process of the camera imaging process. Its research scope is quite extensive, mainly including: target detection and recognition, edge extraction, feature extraction and 3D reconstruction. 3D reconstruction technology is also an image-based modeling technology, which has attracted much attention since its inception. This method only needs two frames of adjacent images to more accurately restore the 3D spatial relationship between the matching feature points in the image and the camera. In this process, the number of matching feature points directly determines the quality of the point cloud ac...

AI Technical Summary

Problems solved by technology

However, it requires the assistance of a depth camera and is very sensitive to the accuracy of the depth map. In a large-scale reconstruction scene, the accuracy of the depth camera is always limited, and the accuracy of the depth camera will directly correlate with the accuracy of the reconstructed point cloud.

[0004] Among the above methods, the stereo vision method requires less calculation, but there are blank areas in the disparity map obtained in the flat area of ​​​​the image texture, so the calculated point cloud density is very low; the motion structure method has high universality , which includes the derivation process from sparse point cloud to dense point cloud, but the density of the obtained dense point cloud still depends on the texture complexity of the image. For images with flat texture, the obtained point cloud density is relatively low; based on depth The image reconstruction method has high precision and does not require the texture complexity of the image. However, this method is highly sensitive to the accuracy of the depth camera and is currently not suitable for 3D reconstruction of large-scale objects.

Images