Method for rebuilding three-dimensional scene of downhole environment

Abstract

The invention provides a method for rebuilding a three-dimensional scene of a downhole environment. The method includes the steps that S1, a depth camera is calibrated to determine internal parameters of the camera; S2, bilinearity filtering is carried out on the depth of a current frame to reduce noise interference; S3, a depth image of the current frame is mapped to a three-dimensional coordinate system of the depth camera; S4, primary registering is carried out through an ICP matching algorithm to obtain a transformation matrix between the depth image of the current frame and a depth image of a previous frame of the depth camera, and the depth image of the current frame can be converted to a global coordinate system through accumulation of the transformation matrixs; S5, the scene is rebuilt through a volume space fusion algorithm. According to the technical scheme, the three-dimensional scene of the downhole environment can be rebuilt economically and rapidly.

Description

technical field [0001] The invention relates to a three-dimensional image processing technology, in particular to a reconstruction method of a large-scale underground scene. Background technique [0002] The use of 3D scene reconstruction technology to create a 3D virtual scene of mine roadway has important practical value for improving coal safety production and miner safety. Traditional 3D reconstruction methods can be divided into two categories according to different ways of obtaining 3D data: 3D reconstruction methods based on laser scanning technology and 3D reconstruction methods based on vision. The laser-based three-dimensional reconstruction method obtains the depth data or distance image of the scene through a laser scanner, and uses the registration of the depth data to realize the alignment of single-frame data and global data; the vision-based three-dimensional reconstruction method uses computer vision methods Three-dimensional model reconstruction refers to ...

AI Technical Summary

Problems solved by technology

At present, the research on 3D reconstruction using Kinect has achieved some research results in the 3D reconstruction of a single object, but it still cannot meet the needs of large-scale underground 3D scene reconstruction.

Images