Improved YOLOv5 for Remote Sensing Image Detection

doi:10.3778/j.issn.1002-8331.2212-0045

Abstract

Abstract: Focusing on that YOLOv5 fails to take into account the issues of poor detection effects, false detection as well as omission caused by complex background information, small detection targets and low percentage of target semantic information in remote sensing image object detection, this paper proposes an improved YOLOv5 for remote sensing target detection. Firstly, a lightweight channel attention block is embedded to the C3 module of feature extraction and fusion module, aiming at enhancing the abilities of local feature extraction and fusion. Secondly, to enhance the multi-scale feature representation capability, a fine-level detection layer that fuses shallow semantic information is added, which helps to detect small targets. Finally, the Copy-Paste data augmentation is leveraged to enrich the diversity of training samples, which further solves the rate problem of high background information and low target area without introducing extra computation cost. Experimental results show that the improved YOLOv5 achieves 0.757 and 0.759 mAP values on the DOTA and DIOR datasets, respectively. It outperforms YOLOv5 by 0.017 and 0.059 gains, as well as obtains obvious accuracy improvements compared with other typical methods, demonstrating the effectiveness of the improved YOLOv5.

Key words: YOLOv5, remote sensing, object detection, attention mechanism, data augmentation

摘要： 针对YOLOv5在遥感图像目标检测中未能考虑到遥感图像背景复杂、检测目标较小且图像中目标语义信息占比过低导致的检测效果不佳和易出现误检漏检等问题，提出了一种改进YOLOv5的遥感图像目标检测方法。将轻量级的通道注意力机制引入到原始YOLOv5的特征提取和特征融合网络的C3模块中，以提升网络局部特征捕获与融合能力；强化对遥感图像的多尺度特征表达能力，通过增加一个融合浅层语义信息的细粒度检测层来提高对小目标的检测效果；使用Copy-Paste数据增强方法来丰富训练样本数量，在不增加模型计算量的情况下可进一步解决遥感图像背景信息占比过高而目标区域占比过低的问题。实验结果表明，改进YOLOv5在公开的DOTA和DIOR遥感图像数据集上mAP结果分别达到0.757和0.759。该方法较原始YOLOv5可提高0.017和0.059，相比于其他典型遥感目标检测方法在精度上也有所提升，证明了改进YOLOv5方法的有效性。

关键词: YOLOv5, 遥感图像, 目标检测, 注意力机制, 数据增强

LIU Tao, DING Xueyan, ZHANG Bingbing, ZHANG Jianxin. Improved YOLOv5 for Remote Sensing Image Detection[J]. Computer Engineering and Applications, 2023, 59(10): 253-261.

刘涛, 丁雪妍, 张冰冰, 张建新. 改进YOLOv5的遥感图像检测方法[J]. 计算机工程与应用, 2023, 59(10): 253-261.

References

[1] ŠEVO I，AVRAMOVIĆ A.Convolutional neural network based automatic object detection on aerial images[J].IEEE Geo-science and Remote Sensing Letters，2016，13（5）：740-744.
[2] FREUND Y，SCHAPIRE R E.Experiments with a new boosting algorithm[C]//13th International Conference on Machine Learning，1996：148-156.
[3] VAPNIK V N.An overview of statistical learning theory[J].IEEE Transactions on Neural Networks，1999，10（5）：988-999.
[4] 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，Salt Lake City，2014：580-587.
[5] GIRSHICK R.Fast R-CNN[C]//Proceedings of the 2015 IEEE International Conference on Computer Vision，Santiago，2015：1440-1448.
[6] 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.
[7] LIU W，ANGUELOV D，ERHAN D，et al.SSD：single shot multibox detector[C]//14th European Conference on Computer Vision.Cham：Springer，2016：21-37.
[8] LIN T Y，GOYAL P，GIRSHICK R，et al.Focal loss for dense object detection[C]//Proceedings of the 2017 IEEE International Conference on Computer Vision，2017：2980-2988.
[9] REDMON J，DIVVALA S，GIRSHICK R，et al.You only look once：unified，real-time object detection[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition，Las Vegas，2016：779-788.
[10] REDMON J，FARHADI A.YOLO9000：better，faster，stronger[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition，Honolulu，2017：6517-6525.
[11] FARHADI A，REDMON J.YOLOv3：an incremental improvement[C]//Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition，2018.
[12] BOCHKOVSKIY A，WANG C Y，LIAO H M，et al.YOLOv4：optimal speed and accuracy of object detection[C]//Proceedings of the 2020 IEEE Conference on Computer Vision and Pattern Recognition，2020：2-7.
[13] 汪鹏，辛雪静，王利琴，等.基于YOLOv3的光学遥感图像目标检测算法[J].激光与光电子学进展，2021，58（20）：20281-20289.
WANG P，XIN X J，WANG L Q，et al.Object detection algorithm of optical remote sensing images based on YOLOv3[J].Laser & Optoelectronics Progress，2021，58（20）：20281-20289.
[14] 赵文清，康怿瑾，赵振兵，等.改进YOLOv5s的遥感图像目标检测[J].智能系统学报，2023，18（1）：86-95.
ZHAO W Q，KANG Y J，ZHAO Z B，et al.A remote sen-
sing image object detection algorithm with improved YOLOv5s[J].CAAI Transactions on Intelligent System，2023，18（1）：86-95.
[15] SHAO J N，YANG Q Y，LUO C Y，et al.Vessel detection from nighttime remote sensing imagery based on deep learning[J].IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing，2021，14：12536-12544.
[16] 李坤亚，欧鸥，刘广滨，等.改进YOLOv5的遥感图像目标检测算法[J/OL].计算机工程与应用[2022-12-14].http：//kns.cnki.net/kcms/detail/11.2127.TP.20221212.1832.022.html.
LI K Y，OU O，LIU G B，et al.Target detection algorithm of remote sensing image based on improved YOLOv5[J/OL].Computer Engineering and Applications[2022-12-14].http：//kns.cnki.net/kcms/detail/11.2127.TP.20221212.1832.
022.html.
[17] 赵文清，孔子旭，周震东，等.增强小目标特征的航空遥感目标检测[J].中国图象图形学报，2021，26（3）：644-653.
ZHAO W Q，KONG Z X，ZHOU Z D，et al.Target detection algorithm of aerial remote sensing based on feature enhancement technology[J].Journal of Image and Graphics，2021，26（3）：644-653.
[18] 林文龙，阿里甫·库尔班，陈一潇，等.面向遥感影像目标检测的ACFEM-RetinaNet算法[J/OL].计算机工程与应用[2022-12-01].http：//kns.cnki.net/kcms/detail/11.2127.TP.
20221125.1132.018.html.
LIN W L，ALIFU K，CHEN Y X，et al.ACFEM-RetinaNet algorithm for remote sensing image target detection[J/OL].Computer Engineering and Applications[2022-12-01].http：//kns.cnki.net/kcms/detail/11.2127.TP.20221125.1132.018.html.
[19] GONG H，MU T，LI Q，et al.Swin-transformer-enabled YOLOv5 with attention mechanism for small object detection on satellite images[J].Remote Sensing，2022，14：2861.
[20] 周华平，郭伟.改进YOLOv5网络在遥感图像目标检测中的应用[J].遥感信息，2022，37（5）：23-30.
ZHOU H P，GUO W.Improved YOLOv5 network in application of remote sensing image object detection[J].Remote Sensing Information，2022，37（5）：23-30.
[21] 李惠惠，范军芳，陈启丽.改进YOLOv5的遥感图像目标检测[J].弹箭与制导学报，2022，42（4）：17-23.
LI H H，FAN J F，CHEN Q L.Improved YOLOv5 remote sensing image target detection[J].Journal of Projectiles，Rockets，Missiles and Guidance，2022，42（4）：17-23.
[22] GOLNAZ G，YIN C，ARAVIND S，et al.Simple Copy-Paste is a strong data augmentation method for instance segmentation[C]//Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition，2021：2918-2928.
[23] LIN T Y，DOLLAR P，GIRSHICK R，et al.Feature pyramid networks for object detection[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition，2017：2117-2125.
[24] LIU S，QI L，QIN H F，et al.Path aggregation network for instance segmentation[C]//Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition，2018：8759-8768.
[25] HU J，SHEN L，SUN G.Squeeze-and-excitation networks[C]//Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition，2018：7132-7141.
[26] DING J，XUE N，XIA S J，et al.Object detection in aerial images：a large-scale benchmark and challenges[J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2021，44（11）：7778-7796.
[27] LI K，WAN G，CHENG G，et al.Object detection in optical remote sensing images：a survey and a new benchmark[J].ISPRS Journal of Photogrammetry and Remote Sensing，2020，159：296-307.