Reconstruction method and device of three-dimensional scene

Abstract

The present invention provides a reconstruction method and device of a three-dimensional scene. The method comprises the following steps that: the first camera attitude of a current frame is obtained according to the depth image of the current frame and the depth image of a reference frame; the second camera attitude of the current frame is estimated according to the gray-scale image of the current frame and the gray-scale image of the reference frame; the fused camera attitude of the current frame is obtained according to the first camera attitude of the current frame and the second camera attitude of the current frame; and a three-dimensional scene model corresponding to the current frame is generated according to the depth image of the current frame and the fused camera attitude of the current frame. According to the reconstruction method of the three-dimensional scene provided by the embodiments of the invention, depth information-based first camera attitude estimation and gray-scale information-based second camera attitude estimation are integrated, so that the cumulative error in the camera attitude estimation can be reduced, and therefore, the reconstruction method can conduct more stably in different scenes, and the reconstructed three-dimensional scene is more accurate.

Description

technical field [0001] Embodiments of the present invention relate to image processing technologies, and in particular, to a method and device for reconstructing a three-dimensional scene. Background technique [0002] Three-dimensional reconstruction refers to the establishment of a mathematical model suitable for computer representation and processing of three-dimensional objects, and it is the key technology to establish a virtual reality that expresses the objective world in the computer. The technology to achieve 3D reconstruction mainly includes Kinect Fusion technology, which is a real-time attitude positioning and matching (SLAM) technology based on color and depth images, that is, the Kinect camera is used to shoot around the 3D object to be reconstructed, so as to reconstruct in real time 3D model of an object. However, the working scene size and resolution of Kinect Fusion are determined by the video memory size of the graphics processor GPU, and the calculation ...

AI Technical Summary

Problems solved by technology

[0004] Although Kintinous solves the problem of limited Kinect Fusion 3D scenes, when the system runs for a long time, since the camera pose in Kintinous is estimated based on the pose trajectory of the previous frame, when the pose trajectory of the camera in a certain frame When an error occurs, it will lead to a cumulative error in the attitude trajectory of the camera in subsequent frames, so the attitude trajectory of the camera deviates from the real trajectory, resulting in inaccurate reconstruction of the 3D scene

Images