基于多层次特征融合和注意力机制的无人机图像小目标检测算法

doi:10.19678/j.issn.1000-3428.0069729

摘要/Abstract

摘要： 无人机(UAV)航拍图像中的目标通常具有尺度密集、易被遮挡且多为小目标等特点,这导致检测过程中容易出现漏检和误检。为应对上述挑战,基于YOLOv5s提出了针对小目标检测的SNA-YOLOv5s算法。首先,引入空间深度转换卷积(SPD-Conv)模块替换原模型的跨步卷积层,避免细节信息丢失,增强小目标特征提取能力;其次,设计新型平均快速空间金字塔池化(AGSPPF)模块,引入平均池化操作缓解池化层在提取特征信息的同时会导致部分信息丢失的问题,提升模型的特征提取能力;再次,新增针对小目标的大尺度检测分支,捕捉浅层特征中丰富的细节信息,提升模型对小目标的检测能力;最后,将归一化注意力机制(NAM)嵌入骨干网络,对特征信息进行加权处理,抑制无效的特征信息。在VisDrone2019数据集和NWPU VHR-10数据集上的训练测试结果表明,该算法的均值平均精度(mAP)分别达到了42.3%和96.5%,与基线模型YOLOv5s相比分别提高了8.4和2.6百分点。通过与其他基于深度学习的主流模型对比实验,进一步验证了该模型的鲁棒性和精确性。

关键词: YOLOv5s模型, 小目标检测, 空间深度转换卷积, 空间金字塔池化, 归一化注意力机制

Abstract: Object detection in Unmanned Aerial Vehicle (UAV) aerial photography images is prone to incorrect or missed detections when the target is small, obstructed, or characterized by dense scales. To address the above challenges, this paper proposes the SNA-YOLOv5s algorithm for small target detection, which is based on YOLOv5s. First, the strided convolution layer in the original model is replaced with the Spatial Depth Transformation Convolution (SPD-Conv) module, eliminating the problem of detail loss caused by strided convolution operations and enhancing the model's ability to extract features from small objects. Second, a novel Average Pyramid Pooling-Fast (AGSPPF) module is designed, and an average pooling operation layer is introduced to address the issue of information loss that occurs while extracting feature information, thereby improving the model's feature extraction capability. Third, a new large-scale detection branch specifically for small targets is added to capture rich details in shallow features and enhance the detection capability for small targets. Finally, the Normalized Attention Mechanism (NAM) is embedded in the backbone network, where feature information is weighted to suppress invalid feature information. The proposed algorithm is trained and tested on the VisDrone2019 and NWPU VHR-10 datasets, on which it achieves mean Average Precision (mAP) of 42.3% and 96.5%, respectively, which is 8.4 and 2.6 percentage points higher than that of the baseline YOLOv5s model. The robustness and accuracy of the proposed model are validated by comparisons with other mainstream deep learning models.

Key words: YOLOv5s model, small target detection, Spatial Depth Transformation Convolution (SPD-Conv), Spatial Pyramid Pooling (SPP), Normalized Attention Mechanism (NAM)

中图分类号:

TP183

张信佳, 王芳. 基于多层次特征融合和注意力机制的无人机图像小目标检测算法[J]. 计算机工程, 2026, 52(2): 148-157.

ZHANG Xinjia, WANG Fang. UAV Image Small Object Detection Algorithm Based on Multi-layer Feature Fusion and Attention Mechanism[J]. Computer Engineering, 2026, 52(2): 148-157.

https://www.ecice06.com/CN/Y2026/V52/I2/148

参考文献

[1] YU W P, YANG T, CHEN C. Towards resolving the challenge of long-tail distribution in UAV images for object detection[C]//Proceedings of the IEEE Winter Conference on Applications of Computer Vision (WACV). Waikoloa, USA: IEEE Press, 2021: 3257-3266.
[2] LIU Y J, YANG F B, HU P. Small-object detection in UAV-captured images via multi-branch parallel feature pyramid networks[J]. IEEE Access, 2020, 8: 145740-145750.
[3] 赵继达, 甄国涌, 储成群. 基于YOLOv8的无人机图像目标检测算法[J]. 计算机工程, 2024, 50(4): 113-120. ZHAO J D, ZHEN G Y, CHU C Q. Unmanned aerial vehicle image target detection algorithm based on YOLOv8[J]. Computer Engineering, 2024, 50(4): 113-120. (in Chinese)
[4] 张静, 农昌瑞, 杨智勇. 基于卷积神经网络的目标检测算法综述[J]. 兵器装备工程学报, 2022, 43(6): 37-47. ZHANG J, NONG C R, YANG Z Y. Review of object detection algorithms based on convolutional neural network[J]. Ordnance Equipment Engineering, 2022, 43(6): 37-47. (in Chinese)
[5] HE K M, ZHANG X Y, RENS S Q, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2014, 37(9): 1904-1916.
[6] GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition. Columbus, USA: IEEE Press, 2014: 580-587.
[7] GIRSHICK R. Fast R-CNN[C]//Proceedings of IEEE International Conference on Computer Vision. Washington D. C., USA: IEEE Press, 2015: 1440-1448.
[8] REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137-1149.
[9] LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot MultiBox detector[C]//Proceedings of ECCV 2016. Berlin, Germany: Springer International Publishing, 2016: 21-37.
[10] LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(2): 318-327.
[11] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, USA: IEEE Press, 2016: 779-788.
[12] REDMON J, FARHADI A. YOLO9000: better, faster,stronger[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, USA: IEEE Press, 2017: 6517-6525.
[13] YE T, QIN W Y, LI Y W, et al. Dense and small object detection in UAV-vision based on a global-local feature enhanced network[J]. IEEE Transactions on Instrumentation and Measurement, 2022, 71: 1-13.
[14] 奉志强, 谢志军, 包正伟, 等. 基于改进YOLOv5的无人机实时密集小目标检测算法[J]. 航空学报, 2023, 44(7): 327106. FENG Z Q,XIE Z J,BAO Z W, et al. Real-time dense small object detection algorithm for UAV based onimproved YOLOv5[J]. Acta Aeronautica et Astronautica Sinica, 2023, 44(7): 327106. (in Chinese).
[15] XU H, ZHENG W L, LIU F X, et al. Unmanned aerial vehicle perspective small target recognition algorithm based on improved YOLOv5[J]. Remote Sensing, 2023, 15(14): 3583.
[16] BAI H Y, YUAN Y M, WU P F, et al. An intelligent detection method for small and weak objects in space[J].Remote Sensing, 2023, 15(12): 3169.
[17] ZHAO L L, ZHU M L. MS-YOLOv7:YOLOv7 based on multi-scale for object detection on UAV aerial photography[J]. Drones, 2023, 7: 188.
[18] 崔丽群, 曹华维. 基于改进YOLOv5的遥感图像目标检测[J]. 计算机工程, 2024, 50(4): 228-236. CUI L Q, CAO H W. Target detection of remote-sensing images based on improved YOLOv5[J]. Computer Engineering, 2024, 50(4): 228-236. (in Chinese)
[19] WANG Y M, ZOU H, YIN M, et al. SMFF-YOLO: a scale-adaptive YOLO algorithm with multi-level feature fusion for object detection in UAV scenes[J]. Remote Sensing, 2023, 15(18): 4580.
[20] MA C J, FU Y Y, WANG D Y, et al. YOLO-UAV: object detection method of unmanned aerial vehicle imagery based on efficient multi-scale feature fusion[J]. IEEE Access, 2023, 11: 126857-126878.
[21] 赵继达, 甄国涌, 储成群. 基于YOLOv8的无人机图像目标检测算法[J]. 计算机工程, 2024, 50(4): 113-120. ZHAO J D, ZHEN G Y, CHU C Q. Unmanned aerial vehicle image target detection algorithm based on YOLOv8[J]. Computer Engineering, 2024, 50(4): 113-120. (in Chinese)
[22] JIAO L C, ZHANG F, LIU F, et al. A survey of deep learning-based object detection[J]. IEEE Access, 2019,7: 128837-128868.
[23] RAJA S, TIE L. No more strided convolutions or pooling: a new CNN building block for low-resolution images and small objects[EB/OL].[2023-12-01]. https://arxiv.org/abs/2208.03641.
[24] LIU Y C, SHAO Z R, TENG Y Y, et al. NAM: normalization-based attention module[EB/OL].[2023-12-01]. https://arxiv.org/abs/2111.12419.
[25] QIU M L, HUANG L, TANG B H. ASFF-YOLOv5: multielement detection method for road traffic in UAV images based on multiscale feature fusion[J]. Remote Sensing, 2022, 14(14): 3498.
[26] ZHU P F, WEN L W, DU D W, et al. Detection and tracking meet drones challenge[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, 44(11): 7380-7399.

选择文件类型/文献管理软件名称

选择包含的内容