[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]

Journal of Projectiles, Rockets, Missiles and Guidance >

2020 , Vol. 40 >Issue 4: 149 - 153

DOI: https://doi.org/10.15892/j.cnki.djzdxb.2020.04.032

A Multi-target Detection Method for Missle-borne Images Based on Improved YOLOv3

  • YANG Chuandong 1, 2 ,
  • LIU Zhen 2 ,
  • MA Hanyu 2 ,
  • XIE Ruichao 1
Expand
  • 1 Army Academy of Artillery and Air Defence Force, Hefei 230031, China
  • 2 High Overload Ammunition Guidance Control and Information Perception Laboratory, Army Academy of Artillery and Air Defence Force, Hefei 230031, China

Received date: 2019-09-07

  Online published: 2025-05-30

Fold

Abstract

Aiming at the characteristics of missile-borne target detection, such as large scale variation, high position accuracy and high real-time requirements, the YOLOv3 method was improved.K-means dimension clustering was applied to the anchor box sizes of different feature maps, which enhanced the detection ability. The position loss function was improved, which contributed to the high position accuracy. Fast NMS algorithm was used to accelerate the prediction process, which improved the real-time performance. The experimental results show that the mAP value of the improved algorithm reaches 93.08% and the frame rate reaches 46.59 fps on the 11 kinds of military target datasets, which was increased by 1.47% and 1.14 fps respectively compared with the original YOLOv3 algorithm, and meets the requirements of missile-borne target detection accuracy and real-time performance.

Key words： missle-borne image; multi-target detection; YOLOv3; position loss; fast NMS

Cite this article

YANG Chuandong , LIU Zhen , MA Hanyu , XIE Ruichao . A Multi-target Detection Method for Missle-borne Images Based on Improved YOLOv3[J]. Journal of Projectiles, Rockets, Missiles and Guidance, 2020 , 40(4) : 149 -153 . DOI: 10.15892/j.cnki.djzdxb.2020.04.032

[an error occurred while processing this directive]

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
何景峰, 冀敏, 李盛, 等. 电视制导导弹目标图像处理改进研究[J]. 现代电子技术, 2017(11): 40-42.

[2]
周志强, 汪渤. 尺度不变特征在自寻的电视制导中的应用[J]. 电光与控制, 2009, 16(9): 6-8.

[3]
杨杰. 图像制导技术及实验研究[D]. 南京: 南京航空航天大学, 2012.

[4]
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 and Machine Intelligence, 2016, 39(6):1137-1149.

[5]
LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2017, 99:2999-3007.

[6]
TIAN Z, SHEN C, CHEN H, et al. FCOS:fully convolutional one-stage object detection[C]// Proceedings of 2019 IEEE/CVF International Conference on Computer Vision(ICCV). [S.l.]: IEEE, 2019:9626-9635.

[7]
鞠默然, 罗海波, 王仲博, 等. 改进的YOLOV3算法及其在小目标检测中的应用[J]. 光学学报, 2019, 39(7):253-260.

[8]
REZATOFIGHI H, TSOI N, GWAK J, et al. Generalized intersection over union:a metric and a loss for bounding box regression[C]// Proceedings of 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR). [S.l.]: IEEE, 2019:658-666.

[9]
LIN T Y, DOLLA P, GIRSHICK R, et al. Feature pyramid networks for objec detection[C]// Proceedings of 30 th IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR). [S.l.]: IEEE, 2017:936-944.

[10]
GIRSHICK R. Fast R-CNN[C]// Proceedings of IEEE International Conference on Computer Vision. [S.l.]: IEEE, 2015:1440-1448.

Options
Outlines

/

〈 〉

[an error occurred while processing this directive]