[1]许小叶,张印辉.基于背景学习的运动工件目标分割方法研究[J].机械与电子,2017,(12):77-80.
　XU Xiaoye,ZHANG Yinhui.Research on Moving Object Segmentation Method Based on Background Learning[J].Machinery & Electronics,2017,(12):77-80.
点击复制

基于背景学习的运动工件目标分割方法研究
分享到：

《机械与电子》[ISSN:1001-2257/CN:52-1052/TH]

卷:
期数:
2017年12期
页码:
77-80
栏目:
智能工程
出版日期:
2017-12-15

文章信息/Info

Title:
Research on Moving Object Segmentation Method Based on Background Learning
文章编号:
1001-2257(2017)12-0077-04
作者:
许小叶; 张印辉
(昆明理工大学机电工程学院,云南 昆明 650500)
Author(s):
XU Xiaoye; ZHANG Yinhui
(Faculty of Mechanical and Electrical Engineering, Kunming University of Science and Technology, Kunming 650500, China)
关键词:
工业检测平台;  机器人视觉;  快速目标分割;  背景学习
Keywords:
industrial detection platform;  robot vision;  fast object segmentation;  background learning
分类号:
TP391
文献标志码:
A
摘要:
精准的目标分割是实现工业机器人对工件目标跟踪和识别的前提和基础。在非限制场景中运动目标快速分割方法的基础上,提出通过高斯混合模型对传送带背景进行学习,实现了工业机器人视觉系统对传送带上运动工件目标的鲁棒分割。实验结果表明,利用该方法对运动目标进行检测,精度比视觉系统中原有的自适应阈值方法提高了7.5%。本文方法适用于具有稳态皮带背景的工业机器人视觉检测平台上的运动目标的高性能分割。
Abstract:
Accurate object segmentation is the premise and foundation for the industrial robot to track and recognize workpiece targets. Based on the method of fast object segmentation of moving target in unrestricted scene, this paper proposes to learn the background of conveyor belt by using Gaussian mixture model to realize the robust segmentation of industrial robot vision system to the moving workpiece on a conveyor belt. The experimental results show that the accuracy of this method is 7.5% higher than that of the original adaptive threshold method in the vision system. The proposed method is applicable for high-performance segmentation of moving object on industrial robot vision testing platform with steady belt background.

参考文献/References:

[1] BARNICH O, VAN Droogenbroeck M. ViBe: A universal background subtraction algorithm for video sequences[J]. IEEE Transactions on Image Processing, 2011, 20(6):1709-1724.
[2] 杨建全, 梁华, 王成友. 视频监控技术的发展与现状[J]. 现代电子技术, 2006, 29(21):84-88.
[3] BROX T, MALIK J. Object segmentation by long term analysis of point trajectories[C]// Computer Vision - ECCV 2010 -, European Conference on Computer Vision, Heraklion, Crete, Greece, September 5-11, 2010, Proceedings. DBLP, 2010:282-295.
[4] PAPAZOGLOU A, FERRARI V. Fast object segmentation in unconstrained video[C]// IEEE International Conference on Computer Vision. IEEE, 2014:1777-1784.
[5] 高斯混合模型--GMM(Gaussian Mixture Model)[EB/OL].http://blog.sina.com.cn/u/1423204580
[6] ACHANTA R, SHAJI A, SMITH K, et al. SLIC superpixels compared to state-of-the-art superpixel methods[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2012, 34(11):2274-2282.
[7] 王森, 伍星, 刘韬,等. 基于二进小波变换的多尺度图切割方法[J]. 计算机工程与应用, 2015, 51(13):9-1.

备注/Memo

备注/Memo:
收稿日期:2017-09-05
基金项目:国家自然科学基金项目(61461022)
作者简介:许小叶(1994-),女,河南南阳人,硕士研究生,研究方向为动态目标分割; 张印辉(1977-),男,陕西西安人,教授,研究方向为图像处理、机器视觉和机器智能。
更新日期/Last Update: 2017-12-25