Low-texture plane scene reconstruction method based on sparse SLAM frame

Abstract

The invention discloses a low-texture plane scene reconstruction method based on a sparse SLAM framework, and the method comprises the following steps: 1, reading an RGB image and a depth image of each frame of image data through an RGB-D camera Kinect; 2, performing the following operations for each acquired image frame: A, obtaining the feature points in the image through the grayscale difference between the pixels in the framework of the sparse SLAM; B, extracting a plane area where sufficient feature points cannot be obtained in the image as a plane road sign; 3, taking the coordinates ofthe feature points in the world coordinate system as map points to perform local map construction; 4, carrying out the bundled adjustment and optimization of a local map and performing the looped detection of the local map; 5, taking the feature points and the plane as road signs for loop detection, and performing the global bundled adjustment to optimize the pose and trajectory of a camera. The method not only can solve the quality problem of low-texture region reconstruction, but also can improve the robustness of a sparse feature point SLAM system.

Description

technical field [0001] The invention relates to the fields of computer vision, digital image processing and simultaneous positioning and mapping, in particular to a low-texture planar scene reconstruction method based on a sparse SLAM framework. Background technique [0002] In the field of computer graphics, 3D reconstruction refers to recovering the shape and position information of objects in 3D space by using information such as depth and texture in 2D images captured by cameras. 3D reconstruction based on visual images has the characteristics of low cost and high degree of automation. [0003] There are many types of vision-based 3D reconstruction according to different classification methods. In recent years, whether based on RGB-D images or 3D reconstruction from other monocular cameras, SLAM frameworks have been widely used. Simultaneous Localization and Mapping (SLAM) problem is to allow localization and reconstruction to be carried out at the same time. Compared ...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problem that the SLAM system based on sparse feature points faces low-texture plane areas, such as walls and cabinets, the reconstruction effect is poor, the tracking is easily lost, and the robustness of the algorithm is not strong enough. A method that integrates plane detection into The reconstruction method in the SLAM system based on feature points can make this type of SLAM system be applied to more life scenes.

Images