A Combined Approach Toward Consistent Reconstructions of Indoor Spaces Based on 6D RGB-D Odometry and KinectFusion

TL;DR

This paper introduces a combined RGB-D odometry and KinectFusion approach for accurate indoor space reconstruction, outperforming existing methods and producing ready-to-use 3D meshes without postprocessing.

Contribution

It integrates 6D RGB-D odometry with KinectFusion to enhance reconstruction accuracy and produce immediate 3D meshes, advancing indoor SLAM techniques.

Findings

Outperforms state-of-the-art RGB-D SLAM accuracy

Produces ready-to-use polygon meshes without postprocessing

Effective on publicly available benchmark datasets

Abstract

We propose a 6D RGB-D odometry approach that finds the relative camera pose between consecutive RGB-D frames by keypoint extraction and feature matching both on the RGB and depth image planes. Furthermore, we feed the estimated pose to the highly accurate KinectFusion algorithm, which uses a fast ICP (Iterative Closest Point) to fine-tune the frame-to-frame relative pose and fuse the depth data into a global implicit surface. We evaluate our method on a publicly available RGB-D SLAM benchmark dataset by Sturm et al. The experimental results show that our proposed reconstruction method solely based on visual odometry and KinectFusion outperforms the state-of-the-art RGB-D SLAM system accuracy. Moreover, our algorithm outputs a ready-to-use polygon mesh (highly suitable for creating 3D virtual worlds) without any postprocessing steps.