面向嵌入式平台的光学遥感飞机目标快速检测算法

秦煜; 陆陈鑫; 刘艳丽; 吴庆学; 张琦; 张小贝

doi:10.11834/jrs.20233285

浏览量 : 0 下载量: 13 CSCD: 0 更多指标

R-PDF
PDF
导出
分享
收藏
专辑

面向嵌入式平台的光学遥感飞机目标快速检测算法
Fast detection algorithm of aircraft targets based on optical remote sensing for embedded platform
2023年页码：1-15
网络出版日期： 2023-11-29 ，
DOI： 10.11834/jrs.20233285

扫描看全文

秦煜，陆陈鑫，刘艳丽，吴庆学，张琦，张小贝.XXXX.面向嵌入式平台的光学遥感飞机目标快速检测算法.遥感学报，XX（XX）： 1-15

QIN Yu，LU Chenxin，LIU Yanli，WU Qingxue，ZHANG Qi，ZHANG Xiaobei. XXXX. Fast detection algorithm of aircraft targets based on optical remote sensing for embedded platform. National Remote Sensing Bulletin， XX（XX）：1-15
秦煜，陆陈鑫，刘艳丽，吴庆学，张琦，张小贝.XXXX.面向嵌入式平台的光学遥感飞机目标快速检测算法.遥感学报，XX（XX）： 1-15 DOI： 10.11834/jrs.20233285.

QIN Yu，LU Chenxin，LIU Yanli，WU Qingxue，ZHANG Qi，ZHANG Xiaobei. XXXX. Fast detection algorithm of aircraft targets based on optical remote sensing for embedded platform. National Remote Sensing Bulletin， XX（XX）：1-15 DOI： 10.11834/jrs.20233285.

摘要

针对目前深度学习检测算法在遥感图像在轨实时处理任务上综合性能不平衡，且在星载嵌入式设备上难以部署的问题，本文基于YOLOX-s算法，提出一种面向嵌入式平台的光学遥感飞机目标快速检测算法LAD-YOLOX（Lightweight Aircraft Detection YOLOX）。首先，在硬件感知网络的设计上，基于ShuffleNetv2设计了超轻量、高精度的骨干模块ES-Block（Enhanced ShuffleNet Block），重构原有的主干特征提取网络。其次，引入GSConv构建轻量级颈部特征融合网络GS-Neck，使得前后端结构的参数量配比均衡，降低了计算复杂度的同时精度损失更小。然后，设计轻量级解耦合检测头网络结构，进一步提升飞机目标分类和定位的特征编码，降低模型参数量，提升检测性能。最后，为LAD-YOLOX算法替换置信度预测损失函数Varifocal Loss及边界框定位损失函数SIoU Loss，提高了模型训练的收敛速度和推理时的精度。仿真实验结果表明，所提算法在RSOD遥感数据集上，检测精度损失极低，计算量压缩为原始YOLOX-s模型的43.72%，检测速度提高24 FPS。在XILINX EK-U1-ZCU102-G评估套件上完成算法部署及加速，对自制飞机数据集的检测速度至少可达到26.53 FPS，能够满足在轨实时、准确检测飞机目标的需求。

Abstract

ObjectiveWith the rapid development of space remote sensing technology

the number of high-resolution optical remote sensing images is increasing exponentially. The detection of high strategic value targets such as aircraft is currently a hot research topic in the field of high-resolution image processing for remote sensing. Traditional remote sensing image object detection algorithms include template matching and traditional machine learning. These algorithms mostly rely on prior knowledge from experts

and the features used are mostly primary manual features limited to pixel level

which have certain limitations and cannot cope with complex and ever-changing backgrounds as well as multimodal and diverse targets. As regards to deep learning technology

remote sensing object detection algorithms include two-stage methods and one-stage methods. Two-stage method has high accuracy

but it consumes a lot of resources and limits processing speed. The YOLO detection algorithms are widely concerned and applied due to their simple network structure

balancing detection accuracy and speed. However

it is difficult for one-stage models to be directly deployed on embedded devices on satellites for real-time detection of aircraft targets

due to the limitations of computing power

storage capacity

and model complexity. Therefore

there is an urgent need to study the lightweight network models which can reduce their demand for computing power and storage. The network models with excellent target detection capabilities are then deployed to aerospace chips with limited resources to complete efficient aircraft target detection tasks.MethodTo solve the problem that it is difficult to deploy current network models with excellent target detection capabilities to aerospace chips with limited computing and storage resources

this paper proposes five targeted designs for the benchmark model which is based on the one-stage YOLOX-s algorithm

and the implementation of the model adopts lightweight design concept. A fast optical remote sensing aircraft target detection algorithm LAD-YOLOX (Lightweight Aircraft Detection YOLOX) has been proposed which is suitable for deployment on embedded platforms. By testing the RSOD dataset and self-made aircraft dataset

the effectiveness of the improved strategy and the generalization of the overall model were verified

achieving extremely low-loss compression of the model. Finally

algorithm deployment and acceleration were completed on the XILINX EK-U1-ZCU102-G evaluation kit

which can meet the requirements of real-time performance and detection accuracy.ResultFirst

during the ablation experiment stage

the effectiveness of each improvement module is verified by gradually accumulating the five improvement strategies which are based on the baseline YOLOX-s. Compared to the YOLOX-s

LAD-YOLOX achieves extremely low-loss model compression. The parameter quantity is reduced to 58.83% of the benchmark

and the calculation quantity is reduced to 43.72% of the benchmark

mAP@.5(%) and mAP@.5:.95(%) only reduced by 0.2%

and the detection speed increased by 24 FPS. Then

the proposed design is applied to the detection task of a self-made aircraft dataset. When the depth and width configuration is set to 0.33 and 0.375

the computational cost is 35.45 GFLOPs

and 54.04% mAP is obtained. FP16 bs=1 is used for inference on RTX4090

the speed can reach 59.50 FPS

which is the best comprehensive performance model in the comparative experiment. Finally

using Quantitative Aware Training

the model was deployed and tested on the ZCU102 evaluation kit to obtain the visualization results and detection speed of aircraft targets. When using 6144×6144 images as input

the average processing speed can reach 26.53 FPS

and the aircraft targets annotated in the self-made aircraft dataset can be accurately identified. Some aircraft targets have overlapping detection frames

misjudgment of difficult cases

and false negatives.ConclusionIn response to the issue of imbalanced comprehensive performance of deep learning detection algorithms in real-time processing of remote sensing images in orbit

and difficulty in deployment on embedded devices on satellites

according to the YOLOX-s algorithm

this paper proposes a fast aircraft detection algorithm named LAD-YOLOX (Lightweight Aircraft Detection YOLOX) which based on optical remote sensing for embedded platforms. First

an ultra-lightweight and high-precision backbone module ES-Block is designed on hardware-aware network to reconstruct the original backbone

which based on ShuffleNetv2. Second

GSConv is introduced to construct lightweight neck which is named GS-Neck. This design equalizes the parameters between the front and the back

reducing the computational complexity and the precision loss. Then

the lightweight decoupling detection head is designed to further improve the feature coding of classification and location

reducing the parameters to improve the detection performance. Finally

Varifocal Loss and SIoU Loss are used in the loss function of LAD-YOLOX

which improves the convergence speed of model training and the precision of reasoning. The simulation results show that the proposed algorithm on the RSOD dataset makes the detection accuracy lossless and the calculation quantity is compressed to 43.72% of the original YOLOX-s model. The detection speed is increased by 24 FPS. The algorithm deployment and acceleration are completed on XILINX EK-U1-ZCU102-G evaluation kit. The detection speed of the self-made aircraft dataset can reach 26.53 FPS

which meets the requirements of real-time and accurate detection of aircraft targets in orbit.

关键词

在轨目标检测轻量化YOLOXRSOD数据集ZCU102

Keywords

In-orbit target detectionlightweightYOLOXRSOD datasetZCU102

references

Lin Y D, He H J, Yin Z K, and Chen F. 2014. Airplane Detection in Optical Remote Sensing Image Based on Sparse-Representation[J]. 43(9):202-207

林煜东, 和红杰, 尹忠科, 陈帆. 2014. 基于稀疏表示的可见光遥感图像飞机检测算法[J]. 光子学报. 43(9):202-207[DOI:10.3788/gzxb20144309.0910001http://dx.doi.org/10.3788/gzxb20144309.0910001]

Lin Y, Zhao M, Pan S D, and An B W. 2019. Method of Aircraft Target Detection in Remote Sensing Images Based on Rotation-invariant Feature[J]. 48(6):159-168

林亿, 赵明, 潘胜达, 安博文. 2019. 基于旋转不变特征的遥感图像飞机目标检测方法[J].光子学报. 48(6):159-168[DOI:10.3788/gzxb20194806.0610002http://dx.doi.org/10.3788/gzxb20194806.0610002]

Sha M M, Li Y, and Li A. 2022. Multiscale aircraft detection in optical remote sensing imagery based on advanced Faster R-CNN. National Remote Sensing Bulletin. 26(8):1624-1635

沙苗苗, 李宇, 李安. 2022. 改进Faster R-CNN的遥感图像多尺度飞机目标检测. 遥感学报26(8):1624-1635[DOI:10.11834/jrs.20219365http://dx.doi.org/10.11834/jrs.20219365]

Cao X, Zou H X, Cheng F, Li R L, He S T. 2021. Aircraft target detection and fine-grained recognition based on RHTC network[J]. Systems Engineering and Electronics. 43(12): 3439-3451

曹旭, 邹焕新, 成飞, 李润林, 贺诗甜. 2021. 基于RHTC网络的飞机目标检测与精细识别[J]. 系统工程与电子技术. 43(12): 3439-3451[DOI:10.12305/j.issn.1001-506X.2021.12.04http://dx.doi.org/10.12305/j.issn.1001-506X.2021.12.04]

Guo Z, Song P, Zhang Y, Yan M L, Sun X, and Sun H. 2018. Aircraft Detection Method Based on Deep Convolutional Neural Network for Remote Sensing Images[J]. Journal of Electronics & Information Technology. 40(11):2684-2690

郭智, 宋萍, 张义, 闫梦龙, 孙显, 孙皓. 2018. 基于深度卷积神经网络的遥感图像飞机目标检测方法[J]. 电子与信息学报. 40(11): 2684-2690[DOI:10.11999/JEIT180117http://dx.doi.org/10.11999/JEIT180117]

Zheng Ge, Liu S T, Wang F, Li Z M, and Sun J. 2021. YOLOX: Exceeding yolo series in2021. [DOI:10.48550/arXiv.2107.08430]

Li C Y, Li L L, Jiang H L, Weng K H, Geng Y F, Li L, Ke Z D, Li Q Y, Cheng M, Nie W Q, Li Y D, Zhang B, Liang Y F, Zhou L Y, Xu X M, Chu X X, Wei X M, and Wei X L.2022. YOLOv6: A Single-Stage Object Detection Framework for Industrial Applications. [DOI:10.48550/arXiv.2209.02976http://dx.doi.org/10.48550/arXiv.2209.02976]

Wang C Y, Bochkovskiy A, and Liao H-Y M. 2022. YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors.[DOI:10.48550/arXiv.2207.02696http://dx.doi.org/10.48550/arXiv.2207.02696]

Ma N N, Zhang X Y, Zheng H T, and Sun J. 2018. ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design. European conference on computer vision. pp.122–138[DOI:10.1007/978-3-030-01264-9_8http://dx.doi.org/10.1007/978-3-030-01264-9_8]

Cohen T, Halfon M, and Schneidman-Duhovny D. 2022. NanoNet: Rapid and Accurate End-to-End Nanobody Modeling by Deep Learning. Frontiers in Immunology. 13:958584[DOI:10.3389/fimmu.2022.958584http://dx.doi.org/10.3389/fimmu.2022.958584]

Tang Y H, Han K, Guo J Y, Xu C, and Wang Y H. 2022. GhostNetV2: Enhance Cheap Operation with Long-Range Attention. [DOI:10.48550/arXiv.2211.12905http://dx.doi.org/10.48550/arXiv.2211.12905]

Fan X N, Yan W, Shi P F and Zhang X W. 2022. Remote sensing image target detection based on a multi-scale deep feature fusion network. National Remote Sensing Bulletin. 26(11): 2292-2303

范新南, 严炜, 史朋飞, 张学武. 2022. 多尺度深度特征融合网络的遥感图像目标检测[J]. 遥感学报. 26(11): 2292-2303[DOI:10.11834/jrs.20210170http://dx.doi.org/10.11834/jrs.20210170]

Wang C Y, Liao H-Y M, Yeh I-H, Wu Y-H, Chen P-Y and Hsieh J-W. 2020. CSPNet: A New Backbone that can Enhance Learning Capability of CNN. IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), pp.1571-1580[DOI:10.1109/CVPRW50498.2020.00203http://dx.doi.org/10.1109/CVPRW50498.2020.00203]

Lin T Y, Dollar P, Girshick R, He K M, Hariharan B, and Belongie S. 2017. Feature Pyramid Networks for Object Detection. IEEE Conference on Computer Vision and Pattern Recognition (CVPR).[DOI:10.1109/CVPR.2017.106http://dx.doi.org/10.1109/CVPR.2017.106]

Li H C, Xiong P F, An J, and Wang L X. 2018. Pyramid Attention Network for Semantic Segmentation. British Machine Vision Conference.[DOI: 10.48550/arXiv.1805.10180http://dx.doi.org/10.48550/arXiv.1805.10180]

Liu Q S, Zhao Z Y, Wang J C, and PI S W. 2022. High Performance YOLOv5: Research on High Performance Target Detection Algorithm for Embedded Platform[J]. Journal of Electronics & Information Technology. 44(0):1-11

刘乔寿, 赵志源, 王均成, 皮胜文. 2022. 高性能YOLOv5：面向嵌入式平台高性能目标检测算法研究[J]. 电子与信息学报. 44(0):1-11[DOI:10.11999/JEIT220413http://dx.doi.org/10.11999/JEIT220413]

Han K, Wang Y H, Tian Q, Guo J Y, Xu C J, and Xu C. GhostNet: More Features From Cheap Operations. 2020. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). pp.1577-1586[DOI:10.1109/cvpr42600.2020.00165http://dx.doi.org/10.1109/cvpr42600.2020.00165]

Hou Q B, Zhou D Q, and Feng J S. Coordinate Attention for Efficient Mobile Network Design. 2021. IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR). Nashville. TN. USA. pp.13708-13717[DOI: 10.1109/CVPR46437.2021.01350http://dx.doi.org/10.1109/CVPR46437.2021.01350]

Lyu C Q, Zhang W W, Huang H A, Zhou Y, Wang Y D, Liu Y Y, Zhang S L, and Chen K. 2023. RTMDet: An Empirical Study of Designing Real-Time Object Detectors. [DOI:10.48550/arXiv.2212.07784http://dx.doi.org/10.48550/arXiv.2212.07784]

Li H L, Li J, Wei H B, Liu Z, Zhan Z F, and Ren Q L. 2022. Slim-neck by GSConv: A better design paradigm of detector architectures for autonomous vehicles.[DOI:10.48550/arXiv.2206.02424http://dx.doi.org/10.48550/arXiv.2206.02424]

Wang H Y, Wang C P, Fu Q, Han Z S and Zhang D D. 2023. Lightweight Ship Detection Based on Optical Remote Sensing Images for Embedded Platform[J]. Acta Optica Sinica. 43(12): 1-22

王慧赢, 王春平, 付强, 韩子硕, 张冬冬. 2023. 面向嵌入式平台的轻量级光学遥感图像舰船检测[J]. 光学学报. 43 (12): 1-22[DOI:10.3788/AOS221689http://dx.doi.org/10.3788/AOS221689]

Zhuang J Y, Qin Z, Yu H, and Chen X C.2023. Task-Specific Context Decoupling for Object Detection.[DOI:10.48550/arXiv.2303.01047http://dx.doi.org/10.48550/arXiv.2303.01047]

Ronneberger O, Fischer P, and Brox T. 2015. U-net: Convolutional networks for biomedical image segmentation. International Conference on Medical image computing and computer-assisted intervention(MICCAI). pp.234-241[DOI:10.1007/978-3-319-24574-4_28http://dx.doi.org/10.1007/978-3-319-24574-4_28]

Gevorgyan Z. 2022. SIoU Loss: More Powerful Learning for Bounding Box Regression.[DOI:10.48550/arXiv.2205.12740http://dx.doi.org/10.48550/arXiv.2205.12740]

Zhang H Y, Wang Y, Dayoub F, and Sünderhauf N. 2021. VarifocalNet: An IoU-aware Dense Object Detector. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR). Nashville. TN. USA. pp.8510-8519 [DOI:10.1109/CVPR46437.2021.00841http://dx.doi.org/10.1109/CVPR46437.2021.00841]

Li K, Wan G, Cheng G, Meng, L Q, and Han J W. 2019. Object Detection in Optical Remote Sensing Images: A Survey and A New Benchmark. ISPRS Journal of Photogrammetry and Remote Sensing. pp. 296–307[DOI:10.1016/j.isprsjprs.2019.11.023http://dx.doi.org/10.1016/j.isprsjprs.2019.11.023]

Li J, ZHOU S, ZHU X C, LI Y and WANG E G. 2022. Remote Sensing Image Aircraft Target Detection Combined with Multiple Channel Attention[J]. Computer Engineering and Applications. 58(1):209-217

李婕, 周顺, 朱鑫潮, 李毅, 王恩果. 2022. 结合多通道注意力的遥感图像飞机目标检测[J]. 计算机工程与应用. 58(1):209-217[DOI: 10.3778/j.issn.1002-8331.2107-0379http://dx.doi.org/10.3778/j.issn.1002-8331.2107-0379]

Gu Y X, Lan X, Fu B Y and Qin X L. 2023. Object detection algorithm for remote sensing images based on geometric adaptation and global perception[J]. Journal of Computer Applications. 43(3):916-922

顾勇翔, 蓝鑫, 伏博毅, 秦小林. 2023. 基于几何适应与全局感知的遥感图像目标检测算法[J]. 计算机应用. 43(3):916-922[DOI: 10.11772/j.issn.1001-9081.2022010071http://dx.doi.org/10.11772/j.issn.1001-9081.2022010071]

Wang C L, Zhao Q, Zhao Y and Guo T. 2022. A Real-Time Remote Sensing Target Detection Algorithm Based on Depth Separable Convolution[J]. Electronics Optics & Control. 29(8):45-49

王成龙, 赵倩, 赵琰, 郭彤. 2022. 基于深度可分离卷积的实时遥感目标检测算法[J]. 电光与控制. 29(8): 45-49[DOI:10.3969/j.issn.1671-637X.2022.08.009http://dx.doi.org/10.3969/j.issn.1671-637X.2022.08.009]

Qin D D, Wan L, He P E, Zhang Y, Guo Y and Chen J. 2022. Multiscale object detection in remote sensing image by combining data fusion and feature selection. National Remote Sensing Bulletin. 26(8):1662-1673

秦登达, 万里, 何佩恩, 张轶, 郭亚, 陈杰. 2022. 结合数据融合与特征选择的遥感影像尺度多样目标检测.遥感学报. 26(8):1662-1673[DOI:10.11834/jrs.20221249http://dx.doi.org/10.11834/jrs.20221249]

Wang X L, Liang Z Y and Liu T. 2023. Feature attention pyramid-based remote sensing image object detection method. National Remote Sensing Bulletin. 27(2):492-501

汪西莉, 梁正印, 刘涛. 2023. 基于特征注意力金字塔的遥感图像目标检测方法.遥感学报. 27(2):492-501[DOI:10.11834/jrs.20235011http://dx.doi.org/10.11834/jrs.20235011]

文章被引用时，请邮件提醒。

提交

暂无数据