Learning Multi-View Stereo With Geometry-Aware Prior
文献类型：期刊论文
| 作者 | Chen, Kehua1; Yuan, Zhenlong1; Xiao, Haihong2; Mao, Tianlu1; Wang, Zhaoqi1 |
| 刊名 | IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
 |
| 出版日期 | 2025-12-01 |
| 卷号 | 35期号:12页码:12396-12409 |
| 关键词 | Feature extraction Accuracy Depth measurement Surface treatment Costs Three-dimensional displays Surface reconstruction Surface texture Image reconstruction Pipelines Multi-view stereo deep learning depth estimation 3D reconstruction |
| ISSN号 | 1051-8215 |
| DOI | 10.1109/TCSVT.2025.3578452 |
| 英文摘要 | Multi-View Stereo (MVS) reconstructs detailed 3D structures from multi-view images by establishing spatial correspondences. While learning-based methods have significantly advanced the MVS task, challenges such as ambiguous matching caused by textureless surfaces and lighting variations persist. To address these issues, we propose GAP-MVSNet, a framework that leverages surface normals from a monocular normal foundation model as priors to enhance the geometric awareness of reconstruction targets. In this work, surface normal priors are seamlessly integrated into the MVS pipeline to improve depth prediction robustness and accuracy. Specifically, we introduce a structure-aware feature pyramid network that incorporates surface normal information and utilizes uncertainty-aware feature resampling to extract robust image features. Additionally, we present the spatial geometry enhanced regularization that combines sampled depth hypotheses with surface normals to generate a spatial geometric prior, guiding the cost regularization process and enforcing strong spatial coherence, particularly in textureless regions. Furthermore, we design a local consistency depth refinement module that utilizes surface normals to establish depth relationships as a local geometric prior, thereby refining classification-based depth predictions and aligning them with ground truth depth. Extensive experiments on the DTU and Tanks & Temples datasets demonstrate that our method achieves state-of-the-art performance. |
| 资助项目 | Strategic Priority Research Program of Chinese Academy of Sciences[XDA0450203] ; Program of National Natural Science Foundation of China[62172392] |
| WOS研究方向 | Engineering |
| 语种 | 英语 |
| WOS记录号 | WOS:001631874000050 |
| 出版者 | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC |
| 源URL | [http://119.78.100.204/handle/2XEOYT63/42977]  |
| 专题 | 中国科学院计算技术研究所 |
| 通讯作者 | Mao, Tianlu |
| 作者单位 | 1.Chinese Acad Sci, Inst Comp Technol, Beijing 100190, Peoples R China 2.South China Univ Technol, Sch Automat Sci & Engn, Guangzhou 511442, Peoples R China |
| 推荐引用方式 GB/T 7714 | Chen, Kehua,Yuan, Zhenlong,Xiao, Haihong,et al. Learning Multi-View Stereo With Geometry-Aware Prior[J]. IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY,2025,35(12):12396-12409. |
| APA | Chen, Kehua,Yuan, Zhenlong,Xiao, Haihong,Mao, Tianlu,&Wang, Zhaoqi.(2025).Learning Multi-View Stereo With Geometry-Aware Prior.IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY,35(12),12396-12409. |
| MLA | Chen, Kehua,et al."Learning Multi-View Stereo With Geometry-Aware Prior".IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY 35.12(2025):12396-12409. |
入库方式： OAI收割
来源：计算技术研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。