Abstract:Three-dimensional reconstruction is a key technology employed to represent virtual reality in the real world, which is valuable in computer vision. Large-scale 3D models have broad application prospects in the fields of smart cities, navigation, virtual tourism, disaster warning, and search-and-rescue missions. Unfortunately, most image-based studies currently prioritize the speed and accuracy of 3D reconstruction in indoor scenes. While there are some studies that address large-scale scenes, there has been a lack of systematic comprehensive efforts to bring together the advancements made in the field of 3D reconstruction in large-scale scenes. Hence, this paper presents a comprehensive overview of a 3D reconstruction technique that utilizes multi-view imagery from large-scale scenes. In this article, a comprehensive summary and analysis of vision-based 3D reconstruction technology for large-scale scenes are presented. The 3D reconstruction algorithms are extensively categorized into traditional and learning-based methods. Furthermore, these methods can be categorized based on whether the sensor actively illuminates objects with light sources, resulting in two categories: active and passive methods. Two active methods, namely, structured light and laser scanning, are briefly introduced. The focus then shifts to structure from motion (SfM), stereo matching, and multi-view stereo (MVS), encompassing both traditional and learning-based approaches. Additionally, a novel approach of neural-radiance-field-based 3D reconstruction is introduced. The workflow and improvements in large-scale scenes are elaborated upon. Subsequently, some well-known datasets and evaluation metrics for various 3D reconstruction tasks are introduced. Lastly, a summary of the challenges encountered in the application of 3D reconstruction technology in large-scale outdoor scenes is provided, along with predictions for future trends in development.

What problem does this paper attempt to address?

The paper primarily focuses on addressing the issue of 3D reconstruction in large-scale environments, particularly in outdoor scenes captured by remote sensing, aviation, and unmanned aerial vehicles. 3D reconstruction technology has a wide range of prospective applications in smart cities, navigation, virtual tourism, disaster early warning, and search and rescue missions. However, most current image-based studies concentrate on the speed and accuracy of 3D reconstruction in indoor settings, while research on large-scale scenes lacks systematic and comprehensive coverage. The authors of the paper provide a thorough review of the techniques for large-scale 3D reconstruction using multi-view imagery. Initially, they categorize 3D reconstruction algorithms into traditional methods and learning-based methods, and further differentiate them into active and passive approaches based on whether the target object is actively illuminated. Active methods include structured light and laser scanning, while passive methods encompass traditional and learning-based approaches based on Structure from Motion (SfM), stereo matching, and Multi-View Stereo (MVS). Additionally, a novel 3D reconstruction method based on neural radiance fields is introduced. The paper elaborates on the workflow of various methods and their improvements in large-scale scenes, and presents commonly used 3D reconstruction datasets and evaluation metrics. Finally, the paper summarizes the challenges faced by the application of 3D reconstruction technology in large-scale outdoor scenes and predicts future development trends. In summary, the paper aims to fill the gap in the field of large-scale scene 3D reconstruction research, providing a comprehensive review to promote the development and application of this field.

Large-Scale 3D Reconstruction from Multi-View Imagery: A Comprehensive Review

Two-Stage Multi-Camera Constrain Mapping Pipeline for Large-Scale 3D Reconstruction

Benchmarking Large-Scale Multi-View 3D Reconstruction Using Realistic Synthetic Images

Efficient 3D reconstruction for urban scenes

A Comprehensive Review of Vision-Based 3D Reconstruction Methods

Outdoor Scene 3D Reconstruction from Multiple Point Cloud

Visual Odometry Based 3D-Reconstruction

3D Reconstruction From Traditional Methods to Deep Learning

DP-MVS: Detail Preserving Multi-View Surface Reconstruction of Large-Scale Scenes

A Cluster-Based 3D Reconstruction System for Large-Scale Scenes

3D reconstruction from spherical images: A review of techniques, applications, and prospects

A review of 3D reconstruction from high-resolution urban satellite images

3D reconstruction of spherical images: a review of techniques, applications, and prospects

Three-Dimensional Dense Reconstruction: A Review of Algorithms and Datasets

3D Reconstruction of Indoor Scenes Via Image Registration

Multi-View Depth Map Sampling for 3D Reconstruction of Natural Scene

Analysis of 3D Reconstruction of Large-Scale Architectural Scene Based on Virtual Reality

3d Reconstruction Based On Stereovision

An End-to-End Approach to Reconstructing 3D Model From Image Set

Deep learning-based 3D reconstruction: a survey

A Review of Techniques for 3D Reconstruction of Indoor Environments