首页 > 论文 > 激光与光电子学进展 > 56卷 > 15期(pp:151202--1)

基于改进Faster R-CNN的子弹外观缺陷检测

Bullet Appearance Defect Detection Based on Improved Faster Region-Convolutional Neural Network

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

摘要

为了实现子弹外观缺陷的自动检测,解决传统机器视觉方法在缺陷检测方面手工设计目标特征耗时和泛化能力差的问题,针对子弹外观缺陷数据集,采用K-means++算法改进锚框的生成方法,提出了Faster R-CNN子弹外观缺陷检测模型。该模型采用卷积神经网络,可以自动提取目标特征,泛化能力强。将该检测模型分别与ZFNet、VGG_CNN_M_1024和VGG16结合,结果表明,与VGG16结合的检测模型的检测精度高于其他两种模型方案,并且在所提算法的基础上,精度提升到了97.75%,速度达到28 frame·s -1。

Abstract

To realize automatic detection of bullet appearance defects and to overcome the limitations associated with traditional machine vision methods, i.e., excessive time required to manually design a target feature and generalization ability is poor in defect detection, we use the K-means++ algorithm to improve the anchor frame generation method and propose a bullet appearance defect detection model based on the improved faster region-convolutional neural network (R-CNN). The proposed model uses a CNN that can automatically extract target features and has strong generalization ability. The detection model is combined with ZFNet, VGG_CNN_M_1024, and VGG16, respectively. Results demonstrate that the detection accuracy of the detection model combined with VGG16 is higher than the others. The results show that that of the improved model demonstrates 97.75% accuracy and the speed reaches 28 frame·s -1.

Newport宣传-MKS新实验室计划
补充资料

DOI:10.3788/LOP56.151202

所属栏目:仪器,测量与计量

收稿日期:2018-12-21

修改稿日期:2019-03-05

网络出版日期:2019-08-01

作者单位    点击查看

马晓云:中国科学院沈阳自动化研究所, 辽宁 沈阳 110016中国科学院机器人与智能制造创新研究院, 辽宁 沈阳 110016中国科学院大学, 北京 100049中国科学院光电信息处理实验室, 辽宁 沈阳 110016辽宁省图像理解与视觉计算重点实验室, 辽宁 沈阳 110016
朱丹:中国科学院沈阳自动化研究所, 辽宁 沈阳 110016中国科学院机器人与智能制造创新研究院, 辽宁 沈阳 110016中国科学院大学, 北京 100049中国科学院光电信息处理实验室, 辽宁 沈阳 110016辽宁省图像理解与视觉计算重点实验室, 辽宁 沈阳 110016
金晨:中国科学院沈阳自动化研究所, 辽宁 沈阳 110016中国科学院机器人与智能制造创新研究院, 辽宁 沈阳 110016中国科学院光电信息处理实验室, 辽宁 沈阳 110016辽宁省图像理解与视觉计算重点实验室, 辽宁 沈阳 110016
佟新鑫:中国科学院沈阳自动化研究所, 辽宁 沈阳 110016中国科学院机器人与智能制造创新研究院, 辽宁 沈阳 110016中国科学院光电信息处理实验室, 辽宁 沈阳 110016辽宁省图像理解与视觉计算重点实验室, 辽宁 沈阳 110016

联系人作者:马晓云(maxiaoyun@sia.cn)

【1】Zhang Y J, Qi X L and Dong C J. Surface detection of gun steel core based on image gray value morphology under machine vision. Journal of Sichuan Ordnance. 33(6), 59-61(2012).
张亚军, 齐杏林, 董成基. 机器视觉下基于图像灰度值形态学的枪弹钢芯表面检测. 四川兵工学报. 33(6), 59-61(2012).

【2】Shi J W, Guo C Y and Liu H N. Study on detection system of bullet surface defect based on machine visionModular Machine Tool & Automatic Manufacturing Technique. 2013(4), 59-64(0).
史进伟, 郭朝勇, 刘红宁. 基于机器视觉的枪弹外观缺陷检测系统研究 组合机床与自动化加工技术. 2013(4), 59-64(0).

【3】Shi J W, Guo C Y, Liu H N et al. Bullet surface defect extraction based on Hough transformation and two-peak algorithm. Fire Control & Command Control. 38(5), 129-132(2013).
史进伟, 郭朝勇, 刘红宁 等. 基于Hough变换和双峰法的枪弹外观缺陷提取. 火力与指挥控制. 38(5), 129-132(2013).

【4】Wang P, Guo C Y and Liu H N. Research on automatic online detection system of bullet surface defect. Journal of Ordnance Engineering College. 27(4), 50-53(2015).
王鹏, 郭朝勇, 刘红宁. 枪弹外观缺陷在线自动检测系统研究. 军械工程学院学报. 27(4), 50-53(2015).

【5】Wang P, Guo C Y and Liu H N. Bullet surface defect recognition and classification based on support vector machine. Computer Engineering & Science. 38(9), 1943-1949(2016).
王鹏, 郭朝勇, 刘红宁. 基于支持向量机的枪弹外观缺陷识别与分类. 计算机工程与科学. 38(9), 1943-1949(2016).

【6】Wang L and Zhang H H. Application of faster R-CNN model in vehicle detection. Journal of Computer Applications. 38(3), 666-670(2018).
王林, 张鹤鹤. Faster R-CNN模型在车辆检测中的应用. 计算机应用. 38(3), 666-670(2018).

【7】Zhang H Y, Wang S N and Hu W B. Improved method for estimating number of people based on convolution neural network. Laser & Optoelectronics Progress. 55(12), (2018).
张红颖, 王赛男, 胡文博. 改进的基于卷积神经网络的人数估计方法. 激光与光电子学进展. 55(12), (2018).

【8】Tian Q, Yuan T Y, Yang D et al. A pedestrian detection method based on dark channel defogging and deep learning. Laser & Optoelectronics Progress. 55(11), (2018).
田青, 袁曈阳, 杨丹 等. 基于暗通道去雾和深度学习的行人检测方法. 激光与光电子学进展. 55(11), (2018).

【9】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: IEEE. 580-587(2014).

【10】Girshick R. Fast R-CNN. [C]∥2015 IEEE International Conference on Computer Vision (ICCV), December 7-13, 2015, Santiago, Chile. New York: IEEE. 1440-1448(2015).

【11】Ren S Q, He K M, Girshick R et al. Faster R-CNN: towards real-time object detection with region proposal networks. IEEE Transactions on Pattern Analysis and Machine Intelligence. 39(6), 1137-1149(2017).

【12】Zeiler M D and Fergus R. Visualizing and understanding convolutional networks. ∥ Fleet D, Pajdla T, Schiele B, et al. Computer vision-ECCV 2014. Lecture notes in computer science. Cham: Springer. 8689, 818-833(2014).

【13】Simonyan K. -04-10)[2018-12-05]. https:∥arxiv. org/abs/1409, (2015).

【14】Li J N and Zhang B H. Face recognition by feature matching fusion combined with improved convolutional neural network. Laser & Optoelectronics Progress. 55(10), (2018).
李佳妮, 张宝华. 特征匹配融合结合改进卷积神经网络的人脸识别. 激光与光电子学进展. 55(10), (2018).

【15】Shelhamer E, Long J and Darrell T. Fully convolutional networks for semantic segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence. 39(4), 640-651(2017).

【16】Everingham M. Eslami S M A, van Gool L, et al. The Pascal visual object classes challenge: a retrospective. International Journal of Computer Vision. 111(1), 98-136(2015).

【17】Jia Y Q, Shelhamer E, Donahue J et al. Caffe. [C]∥Proceedings of the ACM International Conference on Multimedia, November 3-7, 2014, Orlando, Florida, USA. New York: ACM. 675-678(2014).

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

引用该论文

Ma Xiaoyun,Zhu Dan,Jin Chen,Tong Xinxin. Bullet Appearance Defect Detection Based on Improved Faster Region-Convolutional Neural Network[J]. Laser & Optoelectronics Progress, 2019, 56(15): 151202

马晓云,朱丹,金晨,佟新鑫. 基于改进Faster R-CNN的子弹外观缺陷检测[J]. 激光与光电子学进展, 2019, 56(15): 151202

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