River Surface Velocimetry Based on Virtual River Dataset and Modulated-RAFT
In recent years, there has been a surge of interest in image-based velocimetry methods for river surface velocity (RSV) estimation due to their efficiency and accuracy, including large-scale particle image velocimetry (LSPIV), space-time image velocimetry (STIV) and optical flow velocimetry (OFV). A...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2023-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/10103463/ |
_version_ | 1797840079144615936 |
---|---|
author | Yixin Wu Jinbo Zhang Yuqi Cao Zhongyi Wang Guangxin Zhang Dibo Hou |
author_facet | Yixin Wu Jinbo Zhang Yuqi Cao Zhongyi Wang Guangxin Zhang Dibo Hou |
author_sort | Yixin Wu |
collection | DOAJ |
description | In recent years, there has been a surge of interest in image-based velocimetry methods for river surface velocity (RSV) estimation due to their efficiency and accuracy, including large-scale particle image velocimetry (LSPIV), space-time image velocimetry (STIV) and optical flow velocimetry (OFV). Among these methods, OFV-based methods have received significant attention owing to their high field resolution and low tracer requirements. Deep optical flow estimation (DOFE), which is a powerful approach for accurate and efficient estimation of optical flow, has also been employed in OFV-based methods. However, the immeasurability of optical flow often results in the usage of irrelevant datasets for training, leading to limited generalization due to domain drift. Moreover, the high similarity of river surfaces can lead to ambiguous correlation volumes extracted by DOFE models, resulting in mismatches. To address the domain shift and mismatch challenges, we proposed a method to generate optical flow datasets and these datasets are used as the training set for the DOFE model MRAFT, which incorporates correlation volume modulation. Experiment results demonstrate that our method effectively mitigates underlying domain shift and mismatch issues, enabling accurate and robust RSV estimation under velocity ranges of 0-6.0 m/s. Our work facilitates the application of DOFE on RSV estimation and provides optical flow datasets for fine-tuning to other related researches. |
first_indexed | 2024-04-09T16:08:40Z |
format | Article |
id | doaj.art-6d0d724f299c4a93b604277ac3b0fa5f |
institution | Directory Open Access Journal |
issn | 2169-3536 |
language | English |
last_indexed | 2024-04-09T16:08:40Z |
publishDate | 2023-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Access |
spelling | doaj.art-6d0d724f299c4a93b604277ac3b0fa5f2023-04-24T23:00:45ZengIEEEIEEE Access2169-35362023-01-0111382753829010.1109/ACCESS.2023.326763510103463River Surface Velocimetry Based on Virtual River Dataset and Modulated-RAFTYixin Wu0https://orcid.org/0000-0002-3386-277XJinbo Zhang1Yuqi Cao2https://orcid.org/0000-0003-1822-7118Zhongyi Wang3Guangxin Zhang4Dibo Hou5Department of Control Science and Engineering, Zhejiang University, Hangzhou, ChinaDepartment of Control Science and Engineering, Zhejiang University, Hangzhou, ChinaDepartment of Control Science and Engineering, Zhejiang University, Hangzhou, China2City Intelligence, Cloud & AI, Huawei Technologies Company Ltd, Shenzhen, ChinaDepartment of Control Science and Engineering, Zhejiang University, Hangzhou, ChinaDepartment of Control Science and Engineering, Zhejiang University, Hangzhou, ChinaIn recent years, there has been a surge of interest in image-based velocimetry methods for river surface velocity (RSV) estimation due to their efficiency and accuracy, including large-scale particle image velocimetry (LSPIV), space-time image velocimetry (STIV) and optical flow velocimetry (OFV). Among these methods, OFV-based methods have received significant attention owing to their high field resolution and low tracer requirements. Deep optical flow estimation (DOFE), which is a powerful approach for accurate and efficient estimation of optical flow, has also been employed in OFV-based methods. However, the immeasurability of optical flow often results in the usage of irrelevant datasets for training, leading to limited generalization due to domain drift. Moreover, the high similarity of river surfaces can lead to ambiguous correlation volumes extracted by DOFE models, resulting in mismatches. To address the domain shift and mismatch challenges, we proposed a method to generate optical flow datasets and these datasets are used as the training set for the DOFE model MRAFT, which incorporates correlation volume modulation. Experiment results demonstrate that our method effectively mitigates underlying domain shift and mismatch issues, enabling accurate and robust RSV estimation under velocity ranges of 0-6.0 m/s. Our work facilitates the application of DOFE on RSV estimation and provides optical flow datasets for fine-tuning to other related researches.https://ieeexplore.ieee.org/document/10103463/Deep learningoptical flowsurface flow velocityvirtual river datasetscorrelation volume modulation |
spellingShingle | Yixin Wu Jinbo Zhang Yuqi Cao Zhongyi Wang Guangxin Zhang Dibo Hou River Surface Velocimetry Based on Virtual River Dataset and Modulated-RAFT IEEE Access Deep learning optical flow surface flow velocity virtual river datasets correlation volume modulation |
title | River Surface Velocimetry Based on Virtual River Dataset and Modulated-RAFT |
title_full | River Surface Velocimetry Based on Virtual River Dataset and Modulated-RAFT |
title_fullStr | River Surface Velocimetry Based on Virtual River Dataset and Modulated-RAFT |
title_full_unstemmed | River Surface Velocimetry Based on Virtual River Dataset and Modulated-RAFT |
title_short | River Surface Velocimetry Based on Virtual River Dataset and Modulated-RAFT |
title_sort | river surface velocimetry based on virtual river dataset and modulated raft |
topic | Deep learning optical flow surface flow velocity virtual river datasets correlation volume modulation |
url | https://ieeexplore.ieee.org/document/10103463/ |
work_keys_str_mv | AT yixinwu riversurfacevelocimetrybasedonvirtualriverdatasetandmodulatedraft AT jinbozhang riversurfacevelocimetrybasedonvirtualriverdatasetandmodulatedraft AT yuqicao riversurfacevelocimetrybasedonvirtualriverdatasetandmodulatedraft AT zhongyiwang riversurfacevelocimetrybasedonvirtualriverdatasetandmodulatedraft AT guangxinzhang riversurfacevelocimetrybasedonvirtualriverdatasetandmodulatedraft AT dibohou riversurfacevelocimetrybasedonvirtualriverdatasetandmodulatedraft |