Weihua Su
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Weihua Su has a total of 54 co-authors. Here are the top co-authors by number of shared publications:
Dongyuan Zang
4 shared publications- Dongyuan Zang
School of Artificial Intelligence, Hebei University of Technology, Tianjin, China
2 shared publications Huixin Liu
School of Artificial Intelligence, Hebei University of Technology, Tianjin, China
2 shared publications- Jiahe Peng
School of artificial intelligence, Hebei University of Technology, Tianjin 300401, China
2 shared publications - Mianshi Feng
Hebei university of Technology
2 shared publications - Ruixin Li
Tianjin Key Laboratory of Oral and Maxillofacial Function Reconstruction, Tianjin Stomatological Hospital, The Affiliated Stomatological Hospital of Nankai University, Tianjin, China
2 shared publications
Latest Publications
- 29 July 2026
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 23 (4) , 148
STMFusion: a semantic-guided and texture-enhanced Mamba network for multimodal image fusion
- 29 July 2026
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 23 (4) , 148
STMFusion: a semantic-guided and texture-enhanced Mamba network for multimodal image fusion
- 16 October 2025
- Published by Elsevier in Optics & Laser Technology
- Vol. 192
DMDNet: A dual-branch infrared small target detection network based on mamba and difference convolution
- 16 October 2025
- Published by Elsevier in Optics & Laser Technology
- Vol. 192
DMDNet: A dual-branch infrared small target detection network based on mamba and difference convolution
- 18 September 2025
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 22 (5) , 1-20
LMDNet: a lightweight network for infrared small target detection based on mamba and difference convolution
- 18 September 2025
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 22 (5) , 1-20
LMDNet: a lightweight network for infrared small target detection based on mamba and difference convolution
- 19 July 2025
- Published by Elsevier in Optics & Laser Technology
- Vol. 192
ISTDiffuser: Infrared small target image generation via conditional denoising diffusion with contrastive learning
- 19 July 2025
- Published by Elsevier in Optics & Laser Technology
- Vol. 192
ISTDiffuser: Infrared small target image generation via conditional denoising diffusion with contrastive learning
- Open Access
- 1 April 2025
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 2993 (1) , 012082
A precise landing method for drones based on UKF and target position constraints in multi-sensor fusion
To meet the accuracy of navigation and positioning during autonomous landing, this study proposes a multi-sensor fusion path tracking method based on the Unscented Kalman Filter(UKF) and introduces the forced constraint of the target position to enhance navigation accuracy. By building the fusion model of Global Navigation Satellite System (GNSS)/Inertial Navigation System (INS) and visual navigation, UKF is used to handle non-linear problems and integrate multi-source sensor data, and the accuracy of positioning is improved. The study takes the target position as a mandatory constraint and adjusts the state update of UKF to ensure that the UAV maintains high-precision path tracking during landing. Simulation and measured data in open environments, complex terrain, and interference conditions show that this method is significantly better than traditional methods in path tracking accuracy, especially in noise interference conditions. Compared to single-sensor navigation, this method significantly reduces path bias and effectively controls errors.
- 1 April 2025
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 2993 (1) , 012082
Open AccessA precise landing method for drones based on UKF and target position constraints in multi-sensor fusion
To meet the accuracy of navigation and positioning during autonomous landing, this study proposes a multi-sensor fusion path tracking method based on the Unscented Kalman Filter(UKF) and introduces the forced constraint of the target position to enhance navigation accuracy. By building the fusion model of Global Navigation Satellite System (GNSS)/Inertial Navigation System (INS) and visual navigation, UKF is used to handle non-linear problems and integrate multi-source sensor data, and the accuracy of positioning is improved. The study takes the target position as a mandatory constraint and adjusts the state update of UKF to ensure that the UAV maintains high-precision path tracking during landing. Simulation and measured data in open environments, complex terrain, and interference conditions show that this method is significantly better than traditional methods in path tracking accuracy, especially in noise interference conditions. Compared to single-sensor navigation, this method significantly reduces path bias and effectively controls errors.
- 28 March 2025
- Published by Springer Nature
- p. 42-51
Semantic Visual SLAM Algorithm Based on Geometric Constraints
- 28 March 2025
- Published by Springer Nature
- p. 42-51
Semantic Visual SLAM Algorithm Based on Geometric Constraints
- Open Access
- 14 February 2025
- Published by Association for Computing Machinery (ACM)
- p. 353-357
Adaptive terrain travel cost estimation for unmanned vehicles in unstructured environments
- 14 February 2025
- Published by Association for Computing Machinery (ACM)
- p. 353-357
Open AccessAdaptive terrain travel cost estimation for unmanned vehicles in unstructured environments
- 19 September 2024
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 21 (5) , 1-13
High-precision real-time autonomous driving target detection based on YOLOv8
- 19 September 2024
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 21 (5) , 1-13
High-precision real-time autonomous driving target detection based on YOLOv8
- 31 August 2024
- Published by Elsevier in Measurement
- Vol. 240
DCANet: Dense Convolutional Attention Network for infrared small target detection
- 31 August 2024
- Published by Elsevier in Measurement
- Vol. 240
DCANet: Dense Convolutional Attention Network for infrared small target detection
- Open Access
- 24 January 2024
- Published by MDPI AG in Bioengineering
- Vol. 11 (2) , 113
CL-SPO2Net: Contrastive Learning Spatiotemporal Attention Network for Non-Contact Video-Based SpO2 Estimation
Video-based peripheral oxygen saturation (SpO2) estimation, utilizing solely RGB cameras, offers a non-contact approach to measuring blood oxygen levels. Previous studies set a stable and unchanging environment as the premise for non-contact blood oxygen estimation. Additionally, they utilized a small amount of labeled data for system training and learning. However, it is challenging to train optimal model parameters with a small dataset. The accuracy of blood oxygen detection is easily affected by ambient light and subject movement. To address these issues, this paper proposes a contrastive learning spatiotemporal attention network (CL-SPO2Net), an innovative semi-supervised network for video-based SpO2 estimation. Spatiotemporal similarities in remote photoplethysmography (rPPG) signals were found in video segments containing facial or hand regions. Subsequently, integrating deep neural networks with machine learning expertise enabled the estimation of SpO2. The method had good feasibility in the case of small-scale labeled datasets, with the mean absolute error between the camera and the reference pulse oximeter of 0.85% in the stable environment, 1.13% with lighting fluctuations, and 1.20% in the facial rotation situation.
- 24 January 2024
- Published by MDPI AG in Bioengineering
- Vol. 11 (2) , 113
Open AccessCL-SPO2Net: Contrastive Learning Spatiotemporal Attention Network for Non-Contact Video-Based SpO2 Estimation
Video-based peripheral oxygen saturation (SpO2) estimation, utilizing solely RGB cameras, offers a non-contact approach to measuring blood oxygen levels. Previous studies set a stable and unchanging environment as the premise for non-contact blood oxygen estimation. Additionally, they utilized a small amount of labeled data for system training and learning. However, it is challenging to train optimal model parameters with a small dataset. The accuracy of blood oxygen detection is easily affected by ambient light and subject movement. To address these issues, this paper proposes a contrastive learning spatiotemporal attention network (CL-SPO2Net), an innovative semi-supervised network for video-based SpO2 estimation. Spatiotemporal similarities in remote photoplethysmography (rPPG) signals were found in video segments containing facial or hand regions. Subsequently, integrating deep neural networks with machine learning expertise enabled the estimation of SpO2. The method had good feasibility in the case of small-scale labeled datasets, with the mean absolute error between the camera and the reference pulse oximeter of 0.85% in the stable environment, 1.13% with lighting fluctuations, and 1.20% in the facial rotation situation.