Introduction

This is the website for our paper " ChunkFusion: A Learning-based RGB-D 3D Reconstruction Framework via Chunk-wise Integration "

Recent years have witnessed a growing interest in online RGB-D 3D reconstruction. On the premise of ensuring the reconstruction accuracy with noisy depth scans, making the system scalable to various environments is still challenging. In this paper, we devote our efforts to try to fill in this research gap by proposing a scalable and robust RGB-D reconstruction framework, namely ChunkFusion. In ChunkFusion, sparse voxel management is exploited to improve the scalability of online reconstruction. Besides, a chunk-based fusion network is designed to perform a robust integration of the noisy depth measurements on the sparsely allocated voxel chunks. The proposed chunk-based TSDF (truncated signed distance function) integration scheme can accurately restore surfaces with superior visual consistency from noisy depth maps and can guarantee the scalability of online reconstruction simultaneously, making our reconstruction framework widely applicable to scenes with various scales and depth scans with strong noises and outliers. The outstanding scalability and efficacy of our ChunkFusion have been corroborated by extensive experiments.

System Overview

Reconstruction Result

Reconstruction result on ICL-NUIM dataset:

Reconstruction result on self-collected real-world data:

Source Codes

Source code can be found at github

Last update: Oct. 21, 2021