Learning to detect small impurities with superpixel proposals
Guo Y(郭跃)1,2; He YJ(贺一家)1,2; Song HT(宋海涛)2; Yuan K(原魁)2
2017
会议名称IEEE International Conference on Robotics and Biomimetics
会议日期December 5-8, 2017
会议地点Macau SAR, China
摘要
In this paper, we introduce a simplified end-to-end framework for impurity detection in opaque glass bottles with liquor that learns to directly distinguish between small impurities and backgrounds. Despite promising results using convolutional neural networks in various vision tasks, few works have provided specific solutions under inadequate exposures and large background fluctuations. Two contributions are made for this problem. Firstly, we have built a feasible detection system with a cascade hardware structure, and each FPGA provides a host computer with 12 images which are most confident for containing potential impurities respectively. Secondly, most previous convolutional network architectures generally work in large-scale notable object detection benchmarks, however, such networks cannot transfer well when detecting small objects in gray images. Therefore, we propose a superpixel proposal generation method for image augmentation and a fast convolutional network with an overlapped grid structure to detect small impurities, and experiments show that our binary detection results are comparable with human checkers.
关键词Impurity Detection Superpixel Proposal Overlapped Grid Structure Convolutional Neural Network
收录类别EI
文献类型会议论文
条目标识符http://ir.ia.ac.cn/handle/173211/20968
专题智能制造技术与系统研究中心_智能机器人
作者单位1.School of Computer and Control Engineering, University of Chinese Academy of Sciences
2.Institute of Automation, Chinese Academy of Sciences
推荐引用方式
GB/T 7714
Guo Y,He YJ,Song HT,et al. Learning to detect small impurities with superpixel proposals[C],2017.
条目包含的文件 下载所有文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
robio 2017.pdf(1550KB)会议论文 开放获取CC BY-NC-SA浏览 下载
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[Guo Y(郭跃)]的文章
[He YJ(贺一家)]的文章
[Song HT(宋海涛)]的文章
百度学术
百度学术中相似的文章
[Guo Y(郭跃)]的文章
[He YJ(贺一家)]的文章
[Song HT(宋海涛)]的文章
必应学术
必应学术中相似的文章
[Guo Y(郭跃)]的文章
[He YJ(贺一家)]的文章
[Song HT(宋海涛)]的文章
相关权益政策
暂无数据
收藏/分享
文件名: robio 2017.pdf
格式: Adobe PDF
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。