Remote sensing image target detection algorithm based on improved YOLOv8

Abstract

Small size, complicated background, and inaccurate target positioning are characteristics of remote sensing targets in the field of target detection. These factors might impair the detection network’s effectiveness and result in misidentification and omission. A module called LSKB_ECA is designed based on the you only look once version 8 (YOLOv8) backbone detection network. It combines the large selective kernel block (LSKB) with the efficient channel attention (ECA) mechanism to dynamically adjust the spatial sensing field and the localization of the key points, enhancing the model’s attention to the target’s features. A module called C2f_C-DCN is designed to replace the cross stage partial feature fusion (C2f) module in YOLOv8 in order to improve the model’s detection ability in complex scenes. The experimental results demonstrate that the enhanced YOLOv8 algorithm boosts the mean average accuracy by 2.7% and 2.1% on the dataset for object detection in aerial images (DOTA) and detection in optical remote sensing images (DIOR) datasets, respectively, signifying a notable enhancement in performance.

Keywords

A

References

[1] Viola P, Jones M J. Robust real-time face detection. International journal of computer vision, 2004, 57: 137-154 J]

[2] Girshick R, Donahue J, Darrell T, et al. . Rich feature hierarchies for accurate object detection and semantic segmentation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, June 23–28, 2014, Columbus, OH, USA, 2014, New York, IEEE: 580-587 [C]

[3] LI C, LI L, JIANG H, et al. YOLOv6: a single-stage object detection framework for industrial applications[EB/OL]. (2022-09-07) [2025-09-29]. https://arxiv.org/abs/2209.02976.

[4] Wang C Y, Bochkovskiy A, Liao H Y M. YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, June 18–22, 2023, Vancouver, BC, Canada, 2023, New York, IEEE/CVF: 7464-7475 [C]

[5] Lin T Y, Goyal P, Girshick R, et al. . Focal loss for dense object detection. Proceedings of the IEEE International Conference on Computer Vision, October 22–29, 2017, Venice, Italy, 2017, New York, IEEE: 2980-2988 [C]

[6] ZHANG Y, YE M, ZHU G et al. FFCA-YOLO for small object detection in remote sensing images[J]. IEEE transactions on geoscience and remote sensing, 2024, 62.

[7] Xiong X, He M, Li T, et al. . Adaptive feature fusion and improved attention mechanism based small object detection for UAV target tracking. IEEE internet of things journal, 2024, 11(12): 21239-21249 J]

[8] XU X Y, GAO C Y. Improved lightweight infrared vehicle target detection algorithm for YOLOv7-tiny[J]. Journal of computer engineering & applications, 2024, 60(1).

[9] WANG F, WANG H, QIN Z, et al. UAV target detection algorithm based on improved YOLOv8[J]. IEEE access, 2023.

[10] LI Y, HOU Q, ZHENG Z, et al. Large selective kernel network for remote sensing object detection[EB/OL]. (2023-03-16) [2025-09-29]. https://arxiv.org/abs/2303.09030.

[11] Wang Q, Wu B, Zhu P, et al. . ECA-Net: efficient channel attention for deep convolutional neural networks. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, June 14–18, 2020, Seattle, USA, 2020, New York, IEEE/CVF: 11534-11542 [C]

[12] Huang J Q, Fan JF, Lib B. Research on target detection method of CentreNet-based remote sensing image. Journal of ballistic and guidance, 2023, 43(01): 24-31+40 [J]

[13] Dai J, Qi H, Xiong Y, et al. . Deformable convolutional networks. Proceedings of the IEEE International Conference on Computer Vision, October 22–29, 2017, Venice, Italy, 2017, New York, IEEE: 764-773 [C]

[14] Xia G S, Bai X, Ding J, et al. . DOTA: a large-scale dataset for object detection in aerial images. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, June 18–22, 2018, Salt Lake City, UT, USA, 2018, New York, IEEE: 3974-3983 [C]

[15] HE Q, SHEN H. Anchorless multi-scale optical remote sensing image strip target detection method[J]. Advances in lasers and optoelectronics, 2025, 62(4). (in Chinese)

[16] LIAO N S, CAO T X,LIUK Y, et al. Composite feature and multi-scale fusion algorithm for UAV small target detection[J/OL]. Computer engineering and application, 2024: 1–10 [2025-09-29]. http://kns.cnki.net/kcms/detail/11.2127.TP.20241023.1616.006.html.

RIGHTS & PERMISSIONS

Tianjin University of Technology