首页 > 论文 > 激光与光电子学进展 > 57卷 > 22期(pp:221505--1)

基于改进的YOLOv3网络的实时目标检测

Real-Time Object Detection Based on Improved YOLOv3 Network

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

针对YOLOv3算法实时目标检测性能不佳的缺陷,提出了一种适应实时目标检测的改进网络结构以及视频目标检测的新方法。首先,提出的k-means-threshold(k-thresh)方法弥补了k-means算法对聚类中心初始位置十分敏感的问题,在包括三个类别的数据集中进行聚类分析选择合适的锚框;然后,将4倍下采样和8倍下采样特征图拼接融入第三个检测层,以提高对目标的检测精度,将YOLOv3算法的平均准确率均值提高了2%;最后,通过摄像头捕捉图像和前期得到的优秀检测数据来预测新图像的目标以及加入了重新检测阈值,以提高视频检测流畅度。实验结果表明:所提基于改进的YOLOv3网络在检测精度上得以提高,实时性也有所提高,在30 min的实时检测中最大帧率达到64.26 frame/s,相比原始YOLOv3算法,实时检测速度提高了4倍左右。

Abstract

For the shortcoming of the real-time performance of YOLOv3 algorithm in object detection, we propose an improved network structure and a new method for video object detection adapted to real-time object detection. Firstly, the proposed k-means-threshold (k-thresh) method makes up for the problem of its sensitivities to the initial position of the cluster center, and performs cluster analysis on a data set including three categories to select more appropriate anchor boxes. Then, the 4×down-sampling and 8×down-sampling feature maps are stitched together into the third layer detection layer to improve the detection accuracy of the object and increase the the mean average precision of the YOLOv3 algorithm by 2%. Finally, the camera captures the image and the excellent detection data obtained in the previous period to predict the target of the new image and adds a re-detection threshold to improve the smoothness of video detection. The experimental results show that the proposed improved YOLOv3 network improves the detection accuracy and the real-time performance, the maximum frame rate reaches 64.26 frame/s in 30 min of real-time detection, which is 4 times faster than the original YOLOv3 algorithm.

广告组1 - 空间光调制器+DMD
补充资料

中图分类号:TP391.4

DOI:10.3788/LOP57.221505

所属栏目:机器视觉

基金项目:山西省基础研究项目;

收稿日期:2020-04-02

修改稿日期:2020-04-27

网络出版日期:2020-11-01

作者单位    点击查看

孙佳:山西大学物理电子工程学院, 山西 太原 030006
郭大波:山西大学物理电子工程学院, 山西 太原 030006
杨甜甜:山西大学物理电子工程学院, 山西 太原 030006
马识途:山西大学物理电子工程学院, 山西 太原 030006

联系人作者:郭大波(dabo_guo@sxu.edu.cn)

备注:山西省基础研究项目;

【1】CVPR: , 2000.
Schneiderman H, Kanade T, Cat. No. PR00662). June 15-15, 2000, Hilton Head Island, SC, USA. New York: , 2000, 746-751.

【2】Felzenszwalb P. McAllester D, Ramanan D. A discriminatively trained, multiscale, deformable part model [J]. 2008 IEEE Conference on Computer Vision and Pattern Recognition. 2008, 1-8.

【3】Felzenszwalb P F, Girshick R B. McAllester D. Cascade object detection with deformable part models[C]∥2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. June 13-18, 2010, San Francisco, CA, USA. New York: , 2010, 2241-2248.

【4】Forsyth D. Object detection with discriminatively trained part-based models [J]. Computer. 2014, 47(2): 6-7.

【5】Lowe D G. Local feature view clustering for 3D object recognition[C]∥Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001. December 8-14, 2001, Kauai, HI, USA. New York: , 0.

【6】Viola P A, Jones M J . Rapid object detection using a boosted cascade of simple features . [C]∥ Computer Vision and Pattern Recognition. 2001.

【7】Viola P, Jones M J. Robust real-time face detection [J]. International Journal of Computer Vision. 2004, 57(2): 137-154.

【8】Dalal N, Triggs B. Histograms of oriented gradients for human detection[C]∥2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR''''05). June 20-25, 2005, San Diego, CA, USA. New York: , 2005, 886-893.

【9】Ma Q, Zhu B, Cheng Z D, et al. Detection and recognition method of fast low-altitude unmanned aerial vehicle based on dual channel [J]. Acta Optica Sinica. 2019, 39(12): 1210002.
马旗, 朱斌, 程正东, 等. 基于双通道的快速低空无人机检测识别方法 [J]. 光学学报. 2019, 39(12): 1210002.

【10】Girshick R, Donahue J, Darrell T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]∥2014 IEEE Conference on Computer Vision and Pattern Recognition. June 23-28, 2014, Columbus, OH, USA. New York: , 2014, 580-587.

【11】Girshick R. Fast R-CNN [J]. 2015 IEEE International Conference on Computer Vision (ICCV). 2015, 1440-1448.

【12】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.

【13】Feng X Y, Mei W, Hu D S. Aerial target detection based on improved faster R-CNN [J]. Acta Optica Sinica. 2018, 38(6): 0615004.
冯小雨, 梅卫, 胡大帅. 基于改进Faster R-CNN的空中目标检测 [J]. 光学学报. 2018, 38(6): 0615004.

【14】Redmon J, Divvala S, Girshick R, et al. You only look once: , 2016, 779-788.

【15】Redmon J, Farhadi A. YOLO9000: , 2017, 6517-6525.

【16】Redmon J. -04-08)[2020-04-02] . https:∥arxiv. 2018, org/abs/1804: 02767.

【17】Liu W, Anguelov D, Erhan D, et al. SSD: single shot MultiBox detector [J]. Computer Vision - ECCV. 2016, 2016.

【18】Lin T Y, Goyal P, Girshick R, et al. Focal loss for dense object detection[C]∥2017 IEEE International Conference on Computer Vision (ICCV). October 22-29, 2017, Venice, Italy. New York: , 2017, 2999-3007.

【19】Li Y P, Hou L Y, Wang C. Moving objects detection in automatic driving based on YOLOv3 [J]. Computer Engineering and Design. 2019, 40(4): 1139-1144.
李云鹏, 侯凌燕, 王超. 基于YOLOv3的自动驾驶中运动目标检测 [J]. 计算机工程与设计. 2019, 40(4): 1139-1144.

【20】Zou Z X, Shi Z W, Guo Y H, et al. -05-16)[2020-04-02] . https:∥arxiv. 2019, org/abs/1905: 05055.

【21】Ma Q, Zhu B, Zhang H W, et al. Low-altitude UAV detection and recognition method based on optimized YOLOv3 [J]. Laser & Optoelectronics Progress. 2019, 56(20): 201006.
马旗, 朱斌, 张宏伟, 等. 基于优化YOLOv3的低空无人机检测识别方法 [J]. 激光与光电子学进展. 2019, 56(20): 201006.

【22】Ju M R, Luo H B, Wang Z B, et al. Improved YOLO V3 algorithm and its application in small target detection [J]. Acta Optica Sinica. 2019, 39(7): 0715004.
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): 0715004.

【23】Feichtenhofer C, Pinz A, Zisserman A. Detect to track and track to detect[C]∥2017 IEEE International Conference on Computer Vision (ICCV). October 22-29, 2017, Venice, Italy. New York: , 2017, 3057-3065.

【24】Liu R. Video object detection based on deep learning [D]. Guangzhou: South China University of Technology. 2019, 7-8.
刘荣. 基于深度学习的视频目标检测研究 [D]. 广州: 华南理工大学. 2019, 7-8.

【25】Celebi M E, Kingravi H A, Vela P A. A comparative study of efficient initialization methods for the k-means clustering algorithm [J]. Expert Systems with Applications. 2013, 40(1): 200-210.

【26】Lei J S, Jiang T, Wu K, et al. Robust K-means algorithm with automatically splitting and merging clusters and its applications for surveillance data [J]. Multimedia Tools and Applications. 2016, 75(19): 12043-12059.

【27】Lin T Y, Dollár P, Girshick R, et al. -04-19)[2020-04-02] . https:∥arxiv. 2017, org/abs/1612: 03144.

【28】Zhang L, Mou X W. Chinese text clustering algorithm based on Canopy+K-means [J]. Library Tribune. 2018, 38(6): 113-119.
张琳, 牟向伟. 基于Canopy+K-means的中文文本聚类算法 [J]. 图书馆论坛. 2018, 38(6): 113-119.

【29】Kang K, Li H S, Yan J J, et al. T-CNN: tubelets with convolutional neural networks for object detection from videos [J]. IEEE Transactions on Circuits and Systems for Video Technology. 2018, 28(10): 2896-2907.

引用该论文

Sun Jia,Guo Dabo,Yang Tiantian,Ma Shitu. Real-Time Object Detection Based on Improved YOLOv3 Network[J]. Laser & Optoelectronics Progress, 2020, 57(22): 221505

孙佳,郭大波,杨甜甜,马识途. 基于改进的YOLOv3网络的实时目标检测[J]. 激光与光电子学进展, 2020, 57(22): 221505

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF