[1]谭芳喜,肖世德,周亮君,等.基于改进 YOLOv3 算法在道路目标检测中的应用[J].计算机技术与发展,2021,31(08):118-123.[doi:10. 3969 / j. issn. 1673-629X. 2021. 08. 020]
　TAN Fang-xi,XIAO Shi-de,ZHOU Liang-jun,et al.Application in Road Target Detection Based on ImprovedYOLOV3 Algorithm[J].,2021,31(08):118-123.[doi:10. 3969 / j. issn. 1673-629X. 2021. 08. 020]
点击复制

基于改进 YOLOv3 算法在道路目标检测中的应用()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
31
期数:
2021年08期
页码:
118-123
栏目:
应用前沿与综合
出版日期:
2021-08-10

文章信息/Info

Title:
Application in Road Target Detection Based on ImprovedYOLOV3 Algorithm
文章编号:
1673-629X(2021)08-0118-06
作者:
谭芳喜1; 肖世德1; 2; 周亮君1; 李晟尧1
1. 西南交通大学 机械工程学院,四川 成都 610031;
2. 轨道交通运维技术与装备四川省重点实验室,四川 成都 610031
Author(s):
TAN Fang-xi1; XIAO Shi-de1; 2; ZHOU Liang-jun1; LI Sheng-yao1
1. School of Mechanical Engineering,Southwest Jiaotong University,Chengdu 610031,China;
2. Key Laboratory of Sichuan Province for Rail Transit Operation and Maintenance Technology and Equipment,Chengdu 610031,China
关键词:
目标检测; YOLOv3; 深度可分离卷积; SENet; DIoU-NMS 算法
Keywords:
target detection; YOLOv3; deep separable convolution; SENet; DIoU-NMS
分类号:
TP391
DOI:
10. 3969 / j. issn. 1673-629X. 2021. 08. 020
摘要:
为了提高道路环境下目标检测的准确率和实时性,提出一种基于 YOLOv3 的改进检测算法。 通过深度可分离卷积模块减少模型计算量,提高模型的实时性;采用 K-Means++ 聚类算法代替原来的 K-Means 算法生成本数据集所需的先验锚点框,解决 K-Means 算法受初始点选取的影响较大,聚类结果不稳定的问题;在 YOLOv3 的多尺度预测网络中引入SENet(squeeze-and-excitation networks),加强网络对特征的学习能力;改进位置损失函数,解决使用 IoU( intersection overunion)度量时存在无法反映预测框与真实框重合度大小、无法优化 IoU 为零等问题;利用 DIoU-NMS(基于 Distance-IoU 的非极大值抑制)去除冗余框,减少错误抑制,提高检测精度。 实验结果表明,改进算法相对于原算法在检测耗时降低的同时,对 5 类常见目标检测的准确率均有提升。
Abstract:
In order to improve the accuracy and real-time performance of target detection in the road environment,an improved detection algorithm based on YOLOv3 is proposed. The deep separable convolution module is used to reduce the computational load of the model and improve the real-time performance of the model. K-Means++ cluste-ring algorithm is used to replace the original K-Means algorithm to generate the a priori anchor point box,which solves the problem that the K-Means algorithm is greatly affected by the initial point selection and the clustering result is unstable. SENet is introduced into the multi-scale prediction network of YOLOv3 to strengthen the feature learning ability of the network. The location loss function is improved to solve the problems that the intersection over union measurement cannot reflect the intersection degree between the predicted box and the real box,and the IOU cannot be optimized to zero.The DIoU-NMS is used to remove redundant frames,so as to reduce error suppression and improve detection accuracy. The experiment shows that compared with the original algorithm,the proposed algorithm has improved the accuracy of the detection of five types of targets while reducing the detection time.

相似文献/References:

[1]刘晓明 李毓蕙 高燕 郑华强.基于目标区域清晰显示的H.264编码策略[J].计算机技术与发展,2010,(06):29.
　LIU Xiao-ming,LI Yu-hui,GAO Yan,et al.A Coding Strategy of H.264 Based on High-definition Display of Target Region[J].,2010,(08):29.
[2]刘翔 吴谨 祝愿博 康晓晶.基于视频序列的目标检测与跟踪技术研究[J].计算机技术与发展,2009,(11):179.
　LIU Xiang,WU Jin,ZHU Yuan-bo,et al.A Study of Object Detecting and Tracking Based on Video Sequences[J].,2009,(08):179.
[3]曙光 张超 蔡则苏.基于改进的混合高斯模型的目标检测方法[J].计算机技术与发展,2012,(07):60.
　SHU Guang,ZHANG Chao,CAI Ze-su.Target Detection Method Based on Improved Gaussian Mixture Model[J].,2012,(08):60.
[4]刘洁,李目,周少武.一种混沌混合粒子群优化RBF神经网络算法[J].计算机技术与发展,2013,(08):181.
　LIU Jie[],LI Mu[],ZHOU Shao-wu[].An Algorithm of Chaotic Hybrid Particle Swarm Optimization Based on RBF Neural Network[J].,2013,(08):181.
[5]蒋翠清,孙富亮,吴艿芯. 基于相对欧氏距离的背景差值法视频目标检测[J].计算机技术与发展,2015,25(01):37.
　JIANG Cui-qing,SUN Fu-liang,WU Nai-xin. Video Object Detection of Background Subtraction Method Based on Relative Euclidean Distance[J].,2015,25(08):37.
[6]卢官明,衣美佳. 步态识别关键技术研究[J].计算机技术与发展,2015,25(07):100.
　LU Guan-ming,YI Mei-jia. Research on Critical Techniques in Gait Recognition[J].,2015,25(08):100.
[7]高翔,朱婷婷,刘洋. 多摄像头系统的目标检测与跟踪方法研究[J].计算机技术与发展,2015,25(07):221.
　GAO Xiang,ZHU Ting-ting,LIU Yang. Research of Target Detection and Tracking Method for Multi-camera System[J].,2015,25(08):221.
[8]章文洁[][],黄旻[],张桂峰[]. 滤光片多光谱成像中运动目标场景误配准修正[J].计算机技术与发展,2016,26(01):18.
　ZHANG Wen-jie[][],HUANG Min[],ZHANG Gui-feng[]. Misregistration Correction for Moving Object Scene in Filter-type Multispectral Imaging[J].,2016,26(08):18.
[9]施泽浩,赵启军.基于全卷积网络的目标检测算法[J].计算机技术与发展,2018,28(05):55.[doi:10.3969/j.issn.1673－629X.2018.05.013]
　SHI Ze-hao,ZHAO Qi-jun.Object Detection Algorithm Based on Fully Convolutional Neural Network[J].,2018,28(08):55.[doi:10.3969/j.issn.1673－629X.2018.05.013]
[10]张夏清,茅耀斌. 一种改进的ViBe背景提取算法[J].计算机技术与发展,2016,26(07):36.
　ZHANG Xia-qing,MAO Yao-bin. An Improved ViBe Background Generation Method[J].,2016,26(08):36.
[11]李 轩,李 静,王海燕.密集交通场景的目标检测算法研究[J].计算机技术与发展,2020,30(07):46.[doi:10. 3969 / j. issn. 1673-629X. 2020. 07. 011]
　LI Xuan,LI Jing,WANG Hai-yan.Research on Object Detection Algorithm in Dense Traffic Scenes[J].,2020,30(08):46.[doi:10. 3969 / j. issn. 1673-629X. 2020. 07. 011]
[12]余 进,史燕中,王春华,等.一种轻量化目标检测算法研究[J].计算机技术与发展,2020,30(11):42.[doi:10. 3969 / j. issn. 1673-629X. 2020. 11. 008]
　YU Jin,SHI Yan-zhong,WANG Chun-hua,et al.Research of a Lightweight Object Detection Algorithm[J].,2020,30(08):42.[doi:10. 3969 / j. issn. 1673-629X. 2020. 11. 008]
[13]杨学杰,李思毛,李建业,等.面向巡检机器人的电力设备状态检测算法研究[J].计算机技术与发展,2021,31(03):201.[doi:10. 3969 / j. issn. 1673-629X. 2021. 03. 035]
　YANG Xue-jie,LI Si-mao,LI Jian-ye,et al.Research on Algorithm for Equipment Condition Monitoring Based on Inspection Robot[J].,2021,31(08):201.[doi:10. 3969 / j. issn. 1673-629X. 2021. 03. 035]
[14]许 伟,熊卫华.一种改进的轻量级垃圾目标检测算法[J].计算机技术与发展,2022,32(02):63.[doi:10. 3969 / j. issn. 1673-629X. 2022. 02. 010]
　XU Wei,XIONG Wei-hua.An Improved Lightweight Garbage Target Detection Algorithm[J].,2022,32(08):63.[doi:10. 3969 / j. issn. 1673-629X. 2022. 02. 010]
[15]张 剑,王等准,莫光健,等.基于改进 YOLOv3 的高铁异物入侵检测算法[J].计算机技术与发展,2022,32(02):69.[doi:10. 3969 / j. issn. 1673-629X. 2022. 02. 011]
　ZHANG Jian,WANG Deng-zhun,MO Guang-jian,et al.High-speed Rail Foreign Body Intrusion Detection Algorithm Based on Improved YOLOv3[J].,2022,32(08):69.[doi:10. 3969 / j. issn. 1673-629X. 2022. 02. 011]

更新日期/Last Update: 2021-08-10