|Table of Contents|

Remote sensing images vehicle detection based on RDB-YOLOv5(PDF)

长安大学学报（自然科学版）[ISSN:1006-6977/CN:61-1281/TN]

Issue:
2024年3期
Page:
149-160
Research Field:
交通工程
Publishing date:

Info

Title:
Remote sensing images vehicle detection based on RDB-YOLOv5
Author(s):
ZHOU Li1;  HUI Fei1;  ZHANG Jia-yang1;  QI Jian2;  YANG Jing-chao1;  TANG Cui-ren3
(1. School of Electronic Control, Chang'an University, Xi'an 710064, Shaanxi, China; 2. China Construction Eighth Engineering Division CORP., LTD., Xi'an 710001, Shaanxi, China; 3. Xi'an Outer Ring Branch of Shaanxi Transportation Holding Group Co. Ltd, Xi'an 710061, Shaanxi, China)
Keywords:
traffic engineering;  digital image processing;  remote sensing image;  vehicle detection;  rotated bounding box;  dual attention mechanism;  bidirectional feature network
PACS:
U495
DOI:
10.19721/j.cnki.1671-8879.2024.03.013
Abstract:
To solve the problem of dense target and difficult detection of small targetvehicle in remote sensing image, an improved model calledRDB-YOLOv5 based on YOLOv5 was proposed and applied it in remote sensing image vehicle detectionfor the first time. Firstly, to address the problem of arbitrary vehicle orientation in remotesensing images, the existing rotation bounding box-based object detection method CSL(circular smooth label)was improved. Secondly, a multi-scale object detection methodbased on an attention mechanism was proposed to tackle the problem of complex backgroundinformation and reduce detection accuracy due to small vehicle sizes in remote sensingimages. A dual attention mechanism was introduced in the backbone network to combine localand global features, and improvement was made using dilated convolutions. Furthermore,inspired by the idea of bidirectional feature pyramid network, a new shallowfeature and deep feature information transmission paths were added, it was incorporated better tointegrate the positional information of vehicles in shallow layers, and a newdetection head was designed for enhance the detection capability of small target vehicles in the network. The results show that RDB-YOLOv5 achieves a 2.7% increase in mean average precision(mAP)compared to the improved YOLOv5, especially with a 3.5%improvement in small vehicle detection. Compared to traditional models like RCNN, theoverall map is improved by an average of 10%. RDB-YOLOv5 can achieve high detection accuracy on public databases and effectivelysolve the issues of overlap and missed detections caused by horizontal bounding boxdetection in complex scenes of remote sensing images, and the detectionaccuracy of small vehicle targets also improves.8 tabs, 9 figs, 32 refs.

References:

[1] 曹 宝,秦其明,马海建,等.面向对象方法在SPOT5遥感图像分类中的应用:以北京市海淀区为例[J].地理与地理信息科学,2006,22(2):46-49,54.
CAO Bao,QIN Qi-ming,MA Hai-jian,et al.Application of object-oriented approach to SPOT5image classification:A case study in Haidian district,Beijing city[J].Geography andGeo-Information Science,2006,22(2):46-49,54.
[2]吴小波,杨 辽,沈金祥,等.基于背景迭代搜索的高分辨遥感图像汽车检测[J].国土资源遥感,2011,23(4):46-51.
WU Xiao-bo,YANG Liao,SHEN Jin-xiang,et al.Car detection by using high resolution remote sensing image based on background iterative search[J].Remote Sensing for Land &Resources,2011,23(4):46-51.
[3]阳理理,陈雪云,陈家华.基于LFP与RCD(G)特征的遥感图像车辆检测[J].广西大学学报(自然科学版),2018,43(5):1794-1802.
YANG Li-li,CHEN Xue-yun,CHEN Jia-hua.Remote sensing image vehicle detection based on LFP and RCD(G)features[J].Journal of Guangxi University(Natural ScienceEdition),2018,43(5):1794-1802.
[4]王子琦,管振玉,朱轶昇,等.基于改进级联RCNN的遥感图像目标检测[J].计算机工程与设计,2023,44(1):194-202.
WANG Zi-qi,GUAN Zhen-yu,ZHU Yi-sheng,et al.Object detection algorithm of optical remote sensing image based on improved cascade RCNN[J].Computer Engineering andDesign,2023,44(1):194-202.
[5]XIA G S,BAI X A,DING J A,et al.DOTA:A large-scale dataset for object detection in aerial images[C]//IEEE.Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.New York:IEEE,2018:3974-3983.
[6]JIANG Y Y,ZHU X Y,WANG X B,et al.R2CNN:Rotational region CNN for orientation robust scene text detection[J].arXiv:1706.09579,2017.
[7]MA J,SHAO W,YE H,et al.Arbitrary-oriented scene text detection via rotation proposals[J].IEEE Transactions on Multimedia,2018,20(11):3111-3121.
[8]LIU Z K,HU J G,WENG L B,et al.Rotated region based CNN for ship detection[C]//IEEE.Proceedings of 2017 IEEE International Conference on Image Processing(ICIP).Beijing:IEEE,2017:900-904.
[9]LIU Z K,WANG H Z,WENG L B,et al.Ship rotated bounding box space for ship extraction from high-resolution optical satellite images with complex backgrounds[J].IEEE Geoscience and Remote Sensing Letters,2016,13(8):1074-1078.
[10]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.
[11]LIU L,PAN Z X,LEI B.Learning a rotation invariant detector with rotatable bounding box[J].arXiv:1711.09405,2017.
[12]LIN G S,SHEN C H,VAN DEN HENGEL A,et al.Efficient piecewise training of deep structured models for semantic segmentation[C]//IEEE.Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR).New York:IEEE,2016:348.
[13]LIN Z H,FENG M W,DOS SANTOS C N,et al.A structured self-attentive sentence embedding[J].arXiv:1703.03130,2017.
[14]VASWANI A,SHAZEER N,PARMAR N,et al.Attention is all you need[J].arXiv:1706.03762,2017.
[15]李佳琪,邓玉娇,吴湘宁,等.基于注意力及生成对抗网络的遥感影像目标检测[J].计算机系统应用,2022,31(6):182-191.
LI Jia-qi,DENG Yu-jiao,WU Xiang-ning,et al.Object detection in remote sensing image based on attention mechanism and GAN[J].Computer Systems and Applications,2022,31(6):182-191.
[16]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.
[17]ZHANG H,GOODFELLOW I,METAXAS D,et al.Self-attention generative adversarial networks[J].arXiv:1805.08318,2018.
[18]王 红,史金钏,张志伟.基于注意力机制的LSTM的语义关系抽取[J].计算机应用研究,2018,35(5):1417-1420,1440.
WANG Hong,SHI Jin-chuan,ZHANG Zhi-wei.Text semantic relation of LSTM based on attention mechanism[J].Application Research of Computers,2018,35(5):1417-1420,1440.
[19]DAI X Y,CHEN Y P,XIAO B,et al.Dynamic head:Unifying object detection heads with attentions[C]//IEEE.Proceedings of 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR).New York:IEEE,2021:7373-7382.
[20]肖进胜,张舒豪,陈云华,等.双向特征融合与特征选择的遥感影像目标检测[J].电子学报,2022,50(2):267-272.
XIAO Jin-sheng,ZHANG Shu-hao,CHEN Yun-hua,et al.Remote sensing image object detection based on bidirectional feature fusion and feature selection[J].Acta ElectronicaSinica,2022,50(2):267-272.
[21]LIU S,QI L,QIN H F,et al.Path aggregation network for instance segmentation[C]//IEEE.Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.New York:IEEE,2018:8759-8768.
[22]SHU M,ZHONG Y F,LU P Y.Small moving vehicle detection via local enhancement fusion for satellite video[J].International Journal of Remote Sensing,2021,42(19):7189-7214.
[23]ZHAO Q J,SHENG T,WANG Y T,et al.M2Det:A single-shot object detector based on multi-level feature pyramid network[J].Proceedings of the AAAI Conference on ArtificialIntelligence,2019,33(1):9259-9266.
[24]ISLAM M A,ROCHAN M,BRUCE N D B,et al.Gated feedback refinement network for dense image labeling[C]//IEEE.Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition(CVPR).New York:IEEE,2017:4877-4885.
[25]GHIASI G,LIN T Y,LE Q V.NAS-FPN:Learning scalable feature pyramid architecture forobject detection[C]//IEEE.Proceedings of 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR).New York:IEEE,2019:7029-7038.
[26]TAN M X,PANG R M,LE Q V.EfficientDet:Scalable and efficient object detection[C]//IEEE.Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR).New York:IEEE,2020:10781-10790.
[27]张朕通,单玉刚,袁 杰.联合多尺度和注意力机制的遥感影像检测[J].计算机工程与应用,2021,57(9):212-216.
ZHANG Zhen-tong,SHAN Yu-gang,YUAN Jie.Remote sensing image detection algorithmcombining multi-scale and attention mechanism[J].Computer Engineering andApplications,2021,57(9):212-216.
[28]CHEN W Y,ZHAO Y Y,YOU T F,et al.Automatic detection of scattered garbage regions usingsmall unmanned aerial vehicle low-altitude remote sensing images for high-altitude naturalreserve environmental protection[J].Environmental Science & Technology,2021,55(6):3604-3611.
[29]TIAN Z,SHEN C,CHEN H,et al.FCOS:Fully convolutional one-stage object detection[J].arXiv:1904.01355,2019.
[30]LI Y,ZHU J,HOI S C H,et al.Robust estimation of similarity transformation for visual object tracking[J].Proceedings of the AAAI Conference on Artificial Intelligence,2017,33(1):8666-8673.
[31]YANG X,LIU Q,YAN J,et al.R3Det:Refined single-stage detector with feature refinement for rotating object[J].arXiv:1908.05612,2019.
[32]JIANG Y,ZHU X,WANG X,et al.R2CNN:Rotational region CNN for orientation robust scene text detection[J].arXiv:1706.09579,2017.

Memo

Memo:
-
Last Update: 2024-05-01