DY-YOLOv5：基于多重注意力机制的航拍图像目标检测

doi:10.3778/j.issn.1002-8331.2309-0419

摘要/Abstract

摘要： 针对无人机航拍图像中目标小、尺度不一和背景复杂等导致检测精度低的问题，提出一种基于改进YOLOv5的无人机航拍图像目标检测算法DY-YOLOv5。该算法在检测头部分利用具有多重注意力机制的目标检测头方法Dynamic Head，提升检测头在复杂背景下的检测表现。在原模型neck部分增加一次上采样和Concat操作，并执行一个包含极小、小、中目标的多尺度特征检测，提升模型对中、小目标的特征提取能力。引入密集卷积网络DenseNet，将其与YOLOv5s主干网络的C3模块进行融合，提出C3_DenseNet模块，以加强特征传递并预防模型过拟合。在VisDrone2019数据集上应用DY-YOLOv5算法，平均精度均值（mAP）达到了43.9%，较原YOLOv5算法提升了11.4个百分点。召回率（Recall）为41.7%，较原算法提升了9.0个百分点。实验结果证明，改进算法显著提高了无人机航拍图像目标检测的精度。

关键词: 目标检测, YOLOv5, 多重注意力, 密集连接, 多尺度特征

Abstract: Aiming at the problem of low detection accuracy caused by small targets, different scales and complex backgrounds in UAV aerial images, a target detection algorithm for UAV aerial images based on improved YOLOv5 is proposed. The algorithm introduces a target detection head method Dynamic Head with multiple attention mechanisms to replace the original detection head and improves the detection performance of the detection head in complex backgrounds. An upsampling and Concat operation is added to the neck part of the original model, and a multi-scale feature detection including minimal, small and medium targets is performed to improve the feature extraction ability of the model for medium and small targets. DenseNet is introduced and integrated with the C3 module of YOLOv5s backbone network to propose the C3_DenseNet module to enhance feature transfer and prevent model overfitting. The DY-YOLOv5 algorithm is applied to the VisDrone 2019 dataset, and the mean average precision (mAP) reaches 43.9%, which is 11.4 percentage points higher than the original algorithm. The recall rate (Recall) is 41.7%, which is 9.0 percentage points higher than the original algorithm. Experimental results show that the improved algorithm significantly improves the accuracy of target detection in UAV aerial images.

Key words: target detection, YOLOv5, multiple attention, dense connection, multi-scale feature

赵鑫, 陈里里, 杨维川, 张程旺. DY-YOLOv5：基于多重注意力机制的航拍图像目标检测[J]. 计算机工程与应用, 2024, 60(7): 183-191.

ZHAO Xin, CHEN Lili, YANG Weichuan, ZHANG Chengwang. DY-YOLOv5：Target Detection for Aerial Image Based on Multiple Attention[J]. Computer Engineering and Applications, 2024, 60(7): 183-191.

参考文献

[1] LOWE D G. Distinctive image features from scale-invariant key-points[J]. International Journal of Computer Vision, 2004, 60(2) : 91-110.
[2] DALAL N. Histograms of oriented gradients for human detection[C]//2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2005.
[3] GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//2014 IEEE Conference on Computer Vision and Pattern Recognition, 2014.
[4] GIRSHICK R. Fast R-CNN[J]. arXiv:1504.08083, 2015.
[5] REN S, HE K, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2017, 39(6): 1137-1149.
[6] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[J]. arXiv:1506.02640, 2016.
[7] FARHADI A, REDMON J. YOLO9000: better, faster, stronger[J]. arXiv:1612.08242, 2016.
[8] REDMON J, FARHADI A. YOLOv3: an incremental improvement[J]. arXiv:1804.02767, 2018.
[9] BOCHKOVSKIY A, WANG C Y, LIAO H Y M. YOLOv4: optimal speed and accuracy of object detection[J]. ?arXiv:2004.10934, 2020.
[10] BERG A C, FU C Y, SZEGEDY C, et al. SSD: single shot multibox detector[J]. arXiv:1512.02325, 2015.
[11] JAWAHARLALNEHRU A, SAMBANDHAM T, SEKAR V, et al. Target object detection from unmanned aerial vehicle (UAV) images based on improved YOLO algorithm[J]. Electronics, 2022, 11(15): 2343.
[12] 朱瑞鑫, 杨福兴. 运动场景下改进YOLOv5小目标检测算法[J]. 计算机工程与应用, 2023, 59(10): 196-203.
ZHU R X, YANG F X. Improved YOLOv5 small object detection algorithm in motion scenes[J]. Computer Engineering and Applications, 2023, 59(10) : 196-203.
[13] HAN Y Z, HUANG G, SONG S J, et al. Dynamic neural networks: a survey[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, 44(11): 7436-7456.
[14] DAI X Y, CHEN Y P, XIAO B, et al. Dynamic head: unifying object detection heads with attentions[C]//IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021.
[15] HUANG G, LIU Z, LAURENS V D M, et al. Densely con nected convolutional networks[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition, 2017.
[16] LIN T Y, DOLLáR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition, 2017.
[17] SELVARAJU R R, COGSWELL M, DAS A, et al. Grad-CAM: visual explanations from deep networks via gradient-based localization[C]//Proceedings of the IEEE International Conference on Computer Vision, 2017.
[18] HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition, 2016.
[19] WOO S H, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[C]//15th European Conference on Computer Vision (ECCV), Munich, Germany, 2018.
[20] LIU Y, SHAO Z, HOFFMANN N. Global attention mechanism: retain information to enhance channel-spatial interactions[J]. arXiv:2112.05561, 2021.
[21] HU J, SHEN L, ALBANIE S, et al. Squeeze-and-excitation networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(8): 2011-2023.
[22] WANG Q, WU B, ZHU P, et al. ECA-net: efficient channel attention for deep convolutional neural networks[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020.
[23] YANG L X, ZHANG R Y, LI L D, et al. SimAM: a simple, parameter-free attention module for convolutional neural networks[C]//International Conference on Machine Learning, 2021.
[24] CHEN J, KAO S H, HE H, et al. Run, dont walk: chasing higher FLOPS for faster neural networks[C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023.
[25] GAO S H, CHENG M M, ZHAO K, et al. Res2Net: a new multi-scale backbone architecture[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 43(2): 652-662.
[26] CHEN Z, YANG C, LI Q, et al. Disentangle your dense object detector[J]. arXiv:2107.02963, 2021.
[27] ZHOU X, WANG D, KRHENBüHL P. Objects as points[J]. arXiv:1904.07850, 2019.
[28] REIS D, KUPEC J, HONG J, et al. Real-time flying object detection with YOLOv8[J]. arXiv:2305.09972, 2023.
[29] 陈卫彪, 贾小军, 朱响斌, 等. 基于DSM-YOLO v5的无人机航拍图像目标检测[J]. 计算机工程与应用, 2023, 59(18): 226-233.
CHEN W B, JIA X J, ZHU X B, et al. Target detection for UAV image based on DSM-YOLO v5[J]. Computer Engineering and Applications, 2023, 59(18): 226-233.