[an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive]
[an error occurred while processing this directive]

弹箭与制导学报 >

2022 , Vol. 42 >Issue 5: 32 - 37

DOI: https://doi.org/10.15892/j.cnki.djzdxb.2022.05.007

基于改进YOLO_v3-SPP的无人机图像目标检测系统与实现

  • 刘永峰 1 ,
  • 沈延安 2 ,
  • 韦哲 1 ,
  • 李从利 1
展开
  • 1 陆军炮兵防空兵学院兵器工程系,合肥 230031
  • 2 陆军炮兵防空兵学院无人机应用系,合肥 230031

刘永峰(1981—),男,河北邯郸人,副教授,博士,研究方向:计算机视觉。

收稿日期: 2021-09-16

  网络出版日期: 2025-01-16

收起

Target Detection System and Realization of UAV Image Based on Improved YOLO_v3-SPP

  • LIU Yongfeng 1 ,
  • SHEN Yan’an 2 ,
  • WEI Zhe 1 ,
  • LI Congli 1
Expand
  • 1 Department of Weapon Engineering, Army Academy of Artillery and Air Defense, Hefei 230031, China
  • 2 Department of UAV Application, Army Academy of Artillery and Air Defense, Hefei 230031, China

Received date: 2021-09-16

  Online published: 2025-01-16

Fold

摘要

针对现有无人机图像目标检测算法检测率低、误检率高的问题,提出了一种改进YOLO_v3-SPP网络的目标检测算法。以YOLO_v3-SPP网络为基础,在网络结构上进行更多尺度的特征融合,将DarkNet-53中的第3、第4卷积层的信息经下采样后送入网络中进行融合,可提高目标多尺度检测的准确度;引入异常检测网络对目标进行二次判定,可剔除误检样本,有效降低目标检测的误判率。在公开数据集和自建军事目标数据集上进行实验,平均精度提升了4%,改善了现有算法在侦察影像上应用时存在的漏检和误检问题。最后将算法模型移植到硬件平台,系统检测速度小于40 ms,具有较好的检测效果。

关键词： 无人机图像; 目标检测; 特征融合; 异常检测; YOLO_v3-SPP

本文引用格式

刘永峰 , 沈延安 , 韦哲 , 李从利 . 基于改进YOLO_v3-SPP的无人机图像目标检测系统与实现[J]. 弹箭与制导学报, 2022 , 42(5) : 32 -37 . DOI: 10.15892/j.cnki.djzdxb.2022.05.007

Abstract

Aiming at the problems of low detection and high false detection of existing UAV image target detection algorithms, the method of intelligent target detection is carried out. Based on YOLO_v3-SPP network, more scales of feature fusion are performed on the network structure, and the information of the third and fourth convolution layers in DarkNet-53 is down sampled and sent to the network for fusion, which can improve the accuracy of multi-scale target detection; anomaly detection network is introduced to make a second judgment on the target, which can eliminate the misdetected samples and effectively reduce the misjudgment rate of target detection. Experiments on public data sets and self-built military target data sets show that the map value is improved by 4%, which shows that the algorithm improves the problems of missed detection and misdetection in the application of existing algorithms to reconnaissance image. Finally, the algorithm model is transplanted to the hardware platform, and the system detection speed is less than 40 ms, which has a good detection effect.

Key words： UAV image; target detection; feature fusion; anomaly detection; YOLO_v3-SPP

[an error occurred while processing this directive]

参考文献

原文顺序 | 文献年度倒序 | 文中引用次数倒序
[1]
LI Y Z, PANG Y W, CAO J L, et al. Improving single shot object detection with feature scale unmixing[J]. IEEE Transactions on Image Processing, 2021, 30: 2708-2721.

DOI PMID

[2]
FENG M Y, LU H C, YU Y Z. Residual learning for salient object detection[J]. IEEE Transactions on Image Processing, 2020, 29: 4696-4708.

[3]
GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]// IEEE. Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition. New York: IEEE, 2014: 580-587.

[4]
HE K M, ZHANG X Y, REN S Q, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(9): 1904-1916.

DOI PMID

[5]
GIRSHICK R. Fast R-CNN[C]// IEEE. Proceedings of the 2015 IEEE International Conference on Computer Vision. New York: IEEE, 2015: 1440-1448.

[6]
DAI J F, LI Y, HE K M, et al. R-FCN: object detection via region-based fully convolutional networks[C]// NIPS. Proceedings of the 30th Internationl Conference on Neural Information Processing Systems. New York: Curran Associates Inc., 2016: 379-387.

[7]
REDMON J, FARHADI A. YOLO9000: better, faster, stronger[C]// IEEE. Proceedings of the IEEE conference on computer vision and pattern recognition. New York: IEEE, 2017: 7263-7271.

[8]
REDMON J, FARHADI A. Yolov3: an incremental improvement[D]. Washington: University of Washington, 2017.

[9]
LIU W, ANGUELOV D, ERHAN D, et al. Ssd: Single shot multibox detector[C]// CVPR. Proceedings of the European Conference on Computer Vision. Berlin:Springer-Verlag, 2016: 21-37.

[10]
唐志刚. 基于YOLO V3的航拍车辆图像检测方法研究[D]. 赣州: 江西理工大学, 2020.

[11]
郑志强, 刘妍妍, 潘长城, 等. 改进YOLO V3遥感图像飞机识别应用[J]. 电光与控制, 2019, 26(4):28-32.

[12]
鞠默然, 罗海波, 王仲博, 等. 改进的YOLO V3算法及其在小目标检测中的应用[J]. 光学学报, 2019, 39(7):253-260.

[13]
黄梓桐, 阿里甫·库尔班. 无人机平台下的行人与车辆目标实时检测[J]. 计算机工程与应用, 2021, 57(17):169-174.

DOI

[14]
郭智超, 邓建球, 刘爱东, 等. 基于改进SSD的无人机航拍目标检测方法[J]. 兵器装备工程学报, 2021, 42(5):184-190.

[15]
于博文, 吕明. 改进的YOLOv3算法及其在军事目标检测中的应用[J]. 兵工学报, 2022, 43(2):345-354.

DOI

[16]
黄文斌, 陈仁文, 袁婷婷. 改进YOLOv3-SPP的无人机目标检测模型压缩方案[J]. 计算机工程与应用, 2021, 57(21):165-173.

DOI

[17]
LIZNERSKI P, RUFF L, VANDERMEULEN R A, et al. Explainable deep one-class classification[C]// ICLR. Proceedings of the International Conference on Learning Representation. New York: IEEE, 2020: 1760-1785.

[18]
XIA G S, BAI X, DING J, et al. DOTA: a large-scale dataset for object detection in aerial images[C]// IEEE. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. New York: IEEE, 2018: 3974-3983.

Options
文章导航

/

〈 〉

[an error occurred while processing this directive]