Next Article in Journal
The Effects of Decreasing Foot Strike Angle on Lower Extremity Shock Attenuation Measured with Wearable Sensors
Previous Article in Journal
Research on High-Precision Motion Planning of Large Multi-Arm Rock Drilling Robot Based on Multi-Strategy Sampling Rapidly Exploring Random Tree*
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sensors
Volume 25
Issue 9
10.3390/s25092653
sensors-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Hierarchical Motion Field Alignment for Robust Optical Flow Estimation

by
Dianbo Ma
1,*,
Kousuke Imamura
1,
Ziyan Gao
2,
Xiangjie Wang
3 and
Satoshi Yamane
1
1
Graduate School of Natural Science and Technology, Kanazawa University, Kanazawa 9201192, Japan
2
School of Information Science, Japan Advanced Institute of Science and Technology, Nomi 9231292, Japan
3
State Key Laboratory of Automotive Simulation and Control, Jilin University, Changchun 130022, China
*
Author to whom correspondence should be addressed.
Sensors 2025, 25(9), 2653; https://doi.org/10.3390/s25092653
Submission received: 5 March 2025 / Revised: 14 April 2025 / Accepted: 18 April 2025 / Published: 22 April 2025
(This article belongs to the Section Sensing and Imaging)
Browse Figures
Versions Notes

Abstract

Optical flow estimation is a fundamental and long-standing task in computer vision, facilitating the understanding of motion within visual scenes. In this study, we aim to improve optical flow estimation, particularly in challenging scenarios involving small and fast-moving objects. Specifically, we proposed a learning-based model incorporating two key components: the Hierarchical Motion Field Alignment module, which ensures accurate estimation of objects of varying sizes while maintaining manageable computational complexity, and the Correlation Self-Attention module, which effectively handles large displacements, making the model suitable for scenarios with fast-moving objects. Additionally, we introduced a Multi-Scale Correlation Search layer to enhance the four-dimensional cost volume, enabling the model to address various types of motion. Experimental results demonstrate that our model achieves superior generalization performance and significantly improves the estimation of small, fast-moving objects.
Keywords: attention mechanisms; computer vision; correlation; deep learning; image processing; motion estimation; optical flow; recurrent neural networks; residual neural networks; supervised learning attention mechanisms; computer vision; correlation; deep learning; image processing; motion estimation; optical flow; recurrent neural networks; residual neural networks; supervised learning

Share and Cite

MDPI and ACS Style

Ma, D.; Imamura, K.; Gao, Z.; Wang, X.; Yamane, S. Hierarchical Motion Field Alignment for Robust Optical Flow Estimation. Sensors 2025, 25, 2653. https://doi.org/10.3390/s25092653

AMA Style

Ma D, Imamura K, Gao Z, Wang X, Yamane S. Hierarchical Motion Field Alignment for Robust Optical Flow Estimation. Sensors. 2025; 25(9):2653. https://doi.org/10.3390/s25092653

Chicago/Turabian Style

Ma, Dianbo, Kousuke Imamura, Ziyan Gao, Xiangjie Wang, and Satoshi Yamane. 2025. "Hierarchical Motion Field Alignment for Robust Optical Flow Estimation" Sensors 25, no. 9: 2653. https://doi.org/10.3390/s25092653

APA Style

Ma, D., Imamura, K., Gao, Z., Wang, X., & Yamane, S. (2025). Hierarchical Motion Field Alignment for Robust Optical Flow Estimation. Sensors, 25(9), 2653. https://doi.org/10.3390/s25092653

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop