Research on an SSD remote sensing image object detection algorithm based on HSIAM

Chao Chen; Bin Wu; Chao Chen; Bin Wu

doi:10.3934/math.20251010

AIMS Mathematics

2025, Volume 10, Issue 9: 22699-22730. doi: 10.3934/math.20251010

Previous Article Next Article

Research article

Research on an SSD remote sensing image object detection algorithm based on HSIAM

Chao Chen ^1,2,3,
Bin Wu ^{1
,
,}

1.
School of Information and Control Engineering, Southwest University of Science and Technology, Qinglong Avenue 59, Mianyang, P. R. China
2.
Key Laboratory of Numerical Simulation of Sichuan Provincial Universities, Hongqiao Street 1, Neijiang, P. R. China
3.
School of Mathematics and Big Data, Neijiang Normal University, Hongqiao Street 1, Neijiang, P. R. China

Received: 30 June 2025 Revised: 20 August 2025 Accepted: 21 August 2025 Published: 30 September 2025
MSC : 68T05

Aiming at the problems, such as missed detection of small targets, positioning deviation of rotating targets, and complex background interference in remote sensing images, an improved SSD algorithm based on the High-Level Semantic Information Activation Module (HSIAM) and the improved BSWIoU based on Bhattacharyya distance was proposed. The HSIAM module enhances information fusion capabilities within the deep network. The CA mechanism employs adaptive average pooling to enhance focus on central regions of feature maps, distinguishing small targets within complex backgrounds. The RBD_IoU loss function integrates an orientation-matching constraint and a dynamic weighting mechanism to mitigate rotational bounding box regression bias. Experimental results for three benchmark datasets (DIOR, DOTA, and NWPUCHR) showed that, compared with the baseline SSD algorithm, the mAP50 of the improved model increased by approximately 2%. Furthermore, it achieved a balanced trade-off between accuracy and speed, with 12.5% fewer parameters than YOLOv8s. This provides a high-precision and lightweight solution for target detection in remote sensing images.
- SSD,
- Remote sensing object detection,
- HSIAM,
- modified RBD_IoU
Citation: Chao Chen, Bin Wu. Research on an SSD remote sensing image object detection algorithm based on HSIAM[J]. AIMS Mathematics, 2025, 10(9): 22699-22730. doi: 10.3934/math.20251010

Related Papers:

Abstract

Aiming at the problems, such as missed detection of small targets, positioning deviation of rotating targets, and complex background interference in remote sensing images, an improved SSD algorithm based on the High-Level Semantic Information Activation Module (HSIAM) and the improved BSWIoU based on Bhattacharyya distance was proposed. The HSIAM module enhances information fusion capabilities within the deep network. The CA mechanism employs adaptive average pooling to enhance focus on central regions of feature maps, distinguishing small targets within complex backgrounds. The RBD_IoU loss function integrates an orientation-matching constraint and a dynamic weighting mechanism to mitigate rotational bounding box regression bias. Experimental results for three benchmark datasets (DIOR, DOTA, and NWPUCHR) showed that, compared with the baseline SSD algorithm, the mAP50 of the improved model increased by approximately 2%. Furthermore, it achieved a balanced trade-off between accuracy and speed, with 12.5% fewer parameters than YOLOv8s. This provides a high-precision and lightweight solution for target detection in remote sensing images.

References

[1]	K. Ding, Z. Ding, Z. Zhang, M. Yuan, G. Ma, G. Lv, Scd-yolo: A novel object detection method for efficient road crack detection, Multimedia Syst., 30 (2024), 351. http://doi.org/10.1007/s00530-024-01538-y doi: 10.1007/s00530-024-01538-y
[2]	P. Huangfu, L. Dang, A multi-scale pyramid feature fusion-based object detection method for remote sensing images, Int. J. Remote Sens., 44 (2024), 7790–7807. http://doi.org/10.1080/01431161.2023.2288947 doi: 10.1080/01431161.2023.2288947
[3]	L. Xu, Y. Zhao, Y. Zhai, L. Huang, C. Ruan, Small Object Detection in UAV Images Based on YOLOv8n, Int. J. Comput. Intell. Syst., 17 (2024), 223. http://doi.org/10.1007/s44196-024-00632-3 doi: 10.1007/s44196-024-00632-3
[4]	J. Han, D. Zhang, G. Cheng, L. Guo, J. Ren, Object Detection in Optical Remote Sensing Images Based on Weakly Supervised Learning and High-Level Feature Learning, IEEE Trans. Geosci. Remote Sens., 53 (2015), 3325–3337. http://doi.org/10.1109/TGRS.2014.2374218 doi: 10.1109/TGRS.2014.2374218
[5]	V. Zermatten, J. Castillo-Navarro, D. Marcos, D. Tuia, Learning transferable land cover semantics for open vocabulary interactions with remote sensing images, ISPRS J. Photogramm. Remote Sens., 220 (2025), 621–636. http://doi.org/10.1016/j.isprsjprs.2025.01.006 doi: 10.1016/j.isprsjprs.2025.01.006
[6]	Z. Liu, X. Wu, L. Zhang, P. Yu, LightYOLO-S: A lightweight algorithm for detecting small targets, J. Real-Time Image Process., 21 (2024), 111. http://doi.org/10.1007/s11554-024-01485-x doi: 10.1007/s11554-024-01485-x
[7]	H. Wang, H. Qian, S. Feng, Ssd-kdgan: A lightweight SSD target detection method based on knowledge distillation and generative adversarial networks, J. Supercomput., 80 (2024), 23544–23564. http://doi.org/10.1007/s11227-024-06361-w doi: 10.1007/s11227-024-06361-w
[8]	S. Li, F. Yan, Y. Liu, Y. Shen, L. Liu, K. Wang, A multi-scale rotated ship targets detection network for remote sensing images in complex scenarios, Sci. Rep., 15 (2025), 170–183. http://doi.org/10.1038/s41598-025-86601-y doi: 10.1038/s41598-025-86601-y
[9]	X. Yuan, Q. Chen, J. Li, S. Gong, C. Lin, X. Hu, YOLOv5-LC: Enhancing vehicle detection for evening rushing hour, J. Trans. Eng., Part A: Syst., 151 (2025), 4025057. https://doi.org/10.1061/JTEPBS.TEENG-8652 doi: 10.1061/JTEPBS.TEENG-8652
[10]	Z. J. Khow, Y.-F. Tan, H. A. Karim, H. A. A. Rashid, Improved YOLOv8 Model for a Comprehensive Approach to Object Detection and Distance Estimation, IEEE Access, 12 (2024), 63754–63782. http://doi.org/10.1109/ACCESS.2024.3396224 doi: 10.1109/ACCESS.2024.3396224
[11]	N. Singh, C. P. Maurya, B. Mahaur, S. K. Singh, Improved YOLOv11 with weights pruning for road object detection in rainy environment, SIViP, 19 (2025), 473. http://doi.org/10.1007/s11760-025-04070-2 doi: 10.1007/s11760-025-04070-2
[12]	J. Xu, L. Kanokphan, K. Tasaka, Fast and accurate object detection using image cropping/resizing in multi-view 4K sports videos, Proceedings of the 1st International Workshop on Multimedia Content Analysis in Sports (MMSports'18), 2018. https://doi.org/10.1145/3265845.3265852
[13]	C. S. Parvathy, J. P. Jayan, Automatic Lung Cancer Detection Using Computed Tomography Based on Chan Vese Segmentation and SENET, Opt. Mem. Neural Networks, 33 (2024), 339–354. http://doi.org/10.3103/S1060992X2470022X doi: 10.3103/S1060992X2470022X
[14]	Q. Feng, M. Fu, Z. Yao, Y. Liu, T. Liang, Research on small-scale foreign object intrusion detection algorithm for railway tracks based on improved YOLOv8, Modern Electron. Tech., 48 (2025), 174–179. http://doi.org/10.16652/j.issn.1004-373x.2025.11.027 doi: 10.16652/j.issn.1004-373x.2025.11.027
[15]	S. Ding, W. Jing, H. Chen, C. Chen, Yolo Based Defects Detection Algorithm for EL in PV Modules with Focal and Efficient IoU Loss, Appl. Sci., 14 (2024), 7493. http://doi.org/10.3390/app14177493 doi: 10.3390/app14177493
[16]	J. R. Yang, Y. N. Qin, T. X. Li, H. Zhuang, Underground helmet detection algorithm based on improved YOLOv8s, Saf. Coal Mines, 56 (2025), 221–228. http://doi.org/10.13347/j.cnki.mkaq.20241167 doi: 10.13347/j.cnki.mkaq.20241167
[17]	K. Rabia, E. Alperen, Real-time multi-object detection and tracking in UAV systems: Improved YOLOv11-EFAC and optimized tracking algorithms, J. Real-Time Image Proc., 22 (2025), 178. http://doi.org/10.1007/s11554-025-01758-z doi: 10.1007/s11554-025-01758-z
[18]	P. Sharma, I. Malhotra, P. Handa, N. Goel, Real-time detection of household objects using single-shot detection with mobileNet, Artificial Intelligence and Speech Technology 2024, 2025,104–117. http://doi.org/10.1007/978-3-031-91340-2_9
[19]	X. Zhong, CAL-SSD: Lightweight SSD object detection based on coordinated attention, Signal, Image Video Process., 19 (2025), 31. http://doi.org/10.1007/s11760-024-03716-x doi: 10.1007/s11760-024-03716-x
[20]	J. Lei, W. Yang, R. Yang, A Deep Learning Method for Automated Site Recognition of Nasopharyngeal Endoscopic Images, J. Med. Bio. Eng., 45 (2025), 240–251. http://doi.org/10.1007/s40846-025-00936-5 doi: 10.1007/s40846-025-00936-5
[21]	Y. Hou, Y. Rao, H. Song, H. Song, Z. Nie, T. Wang, et al., A Rapid Detection Method for Wheat Seedling Leaf Number in Complex Field Scenarios Based on Improved YOLOv8, Smart Agric., 6 (2024), 128–137. http://doi.org/10.12133/j.smartag.SA202403019 doi: 10.12133/j.smartag.SA202403019
[22]	J. Li, Q. Hou, J. Xing, J. Ju, SSD object detection model based on multi-frequency feature theory, IEEE Access, 8 (2020), 82294–82305. http://doi.org/10.1109/ACCESS.2020.2990477 doi: 10.1109/ACCESS.2020.2990477
[23]	W. Zheng, W. Tang, S. Chen, L. Jiang, C. Fu, CIA-SSD: Confident iou-aware single-stage object detector from point cloud, The Thirty-Fifth AAAI Conference on Artificial Intelligence (AAAI-21), 35 (2021), 3555–3562. https://doi.org/10.1609/aaai.v35i4.16470 doi: 10.1609/aaai.v35i4.16470
[24]	L. Gong, X. Huang, Y. Chao, J. Chen, B. Lei, An enhanced SSD with feature cross-reinforcement for small-object detection, Appl. Intell., 53 (2023), 19449–19465. doilinkhttps://doi.org/10.1007/s10489-023-04544-1 doi: 10.1007/s10489-023-04544-1
[25]	L. Lin, H. Zhao, S. Gao, J. Wang, Z. Zhang, Spatial-Spectral Linear Extrapolation for Cross-Scene Hyperspectral Image Classification, Remote Sens., 17 (2025), 1816. http://doi.org/10.3390/rs17111816 doi: 10.3390/rs17111816
[26]	F. Guo, Z. Li, G. Ren, L. Wang, J. Zhang, J. Wang, Instance-Wise Domain Generalization for Cross-Scene Wetland Classification With Hyperspectral and LiDAR Data, IEEE Trans. Geosci. Remote Sens., 63 (2025). http://doi.org/10.1109/TGRS.2024.3519900
[27]	A. Sarkar, U. Nandi, B. Paul, S. Kr. Ghosal, M. M. Singh, J. K. Mandal, et al., Searching Optimizers for Deep Learning Based Hyperspectral Image Classification, Computational Technologies and Electronics. ICCTE 2023, 2025. https://doi.org/10.1007/978-3-031-81935-3_7
[28]	W. W. Y. Ng, Q. Zhang, C. Zhong, J. Zhang, Improving domain generalization by hybrid domain attention and localized maximum sensitivity, Neural Networks, 171 (2024), 320–331. http://doi.org/10.1016/j.neunet.2023.12.014 doi: 10.1016/j.neunet.2023.12.014
[29]	H. Zhou, A. Liu, C. Zhang, P. Zhu, Q. Zhang, M. Kankanhalli, Multi-Modal Meta-Transfer Fusion Network for Few-Shot 3D Model Classification, Int. J. Comput. Vis., 132 (2024), 673–688. http://doi.org/10.1007/s11263-023-01905-8 doi: 10.1007/s11263-023-01905-8
[30]	Z. Zhang, D. Gao, D. Liu, G. Shi, Spectral-Spatial Domain Attention Network for Hyperspectral Image Few-Shot Classification, Remote Sens., 16 (2024), 22–35. http://doi.org/10.3390/rs16030592 doi: 10.3390/rs16030592

Reader Comments

Your name:*

Email:*
© 2025 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)