Three-dimensional reconstruction device and method based on multi-view structure

Abstract

The invention discloses a three-dimensional reconstruction device and method based on a multi-view structure. Firstly, required data are obtained through an independently designed three-dimensional reconstruction device based on a multi-view structure, and panoramic scanning and pose estimation among multiple cameras are achieved so that subsequent indoor three-dimensional scene reconstruction or panoramic image generation can be carried out. In addition, a vision-wheel speed meter tight coupling frame is provided by fusing information of a plurality of sensors and utilizing a factor graph optimization mode so as to realize high-precision and robust state estimation and reconstruction. According to the method, the consistency of partial motion of the camera is limited through the three-dimensional reconstruction device so as to restrain the degree of freedom of the camera, and the motion of the camera is more accurately estimated through the provided tight coupling frame in a factor graph optimization mode, so that the reconstruction result is more accurate.

Description

technical field [0001] The present invention relates to the field of indoor three-dimensional reconstruction and the field of multi-purpose structures, and in particular relates to a method for performing three-dimensional reconstruction of indoor scenes by way of multi-purpose structures. Background technique [0002] With the vigorous development of AI technology and the continuous emergence of new equipment, 3D reconstruction has become a hot research topic in the field of computer graphics. Its main task is to use multi-view geometry, probability statistics and optimization theory based on data collected by various sensors. 3D modeling of the real physical world and the establishment of a bridge between the real world and the virtual world. Therefore, 3D reconstruction has a wide range of applications in many different fields such as manufacturing, medical treatment, film and television production, cultural relic protection, augmented reality, virtual reality, and positi...

AI Technical Summary

Problems solved by technology

While it is possible to recover coarse depth information by using 2D RGB cameras for canonical stitching and VR / AR applications, the depth quality is usually not acceptable for high-definition 3D reconstruction

Current 3D-to-3D technology based on a single RGB-D camera limits the DOF field of view when the sensor is in motion, and thus cannot cover most spherical panoramas

This narrow field of view can be resolved by utilizing multiple RGB-D cameras (e.g. arranged vertically for horizontal rotation), but this introduces new camera calibration and synchronization issues

[0004] In addition, the traditional indoor 3D reconstruction or panoramic scanning scans the indoor scene frame by frame through a single depth camera held by a person. The problem with this method is that it is very inefficient, and it is difficult to control the speed of movement during the scanning process. with noise

Images