[1]张宗彬.改进的混合高斯模型的运动对象分割算法[J].应用科技,2010,37(05):33-36.[doi:10.3969/j.issn.1009-671X.2010.05.008]
 ZHANG Zong-bin.A moving objects segmentation algorithm based on improved GMM[J].Applied science and technology,2010,37(05):33-36.[doi:10.3969/j.issn.1009-671X.2010.05.008]
点击复制

改进的混合高斯模型的运动对象分割算法(/HTML)
分享到:

《应用科技》[ISSN:1009-671X/CN:23-1191/U]

卷:
第37卷
期数:
2010年05期
页码:
33-36
栏目:
现代电子技术
出版日期:
2010-05-05

文章信息/Info

Title:
A moving objects segmentation algorithm based on improved GMM
文章编号:
1009-671X (2010) 05-0033-04
作者:
张宗彬
哈尔滨工程大学 信息与通信工程学院,黑龙江 哈尔滨 150001
Author(s):
ZHANG Zong-bin
College of Information and Communication Engineering, Harbin Engineering University, Harbin 150001, China)
关键词:
运动检测混合高斯模型鬼影光线突变阴影消除
Keywords:
moving objects detection GMM ghost sudden illumination change shadow removal
分类号:
TP393
DOI:
10.3969/j.issn.1009-671X.2010.05.008
文献标志码:
A
摘要:
针对视频序列中运动对象分割问题,提出一种改进的混合高斯模型分割算法.该算法首先由混合高斯模型得到前景,之后用当前帧的前景区域与上一帧对应位置做差,区分出实际变化区域及误检区域并为误检区域赋予较大的更新速率,从而有效地改善了长时间静止物体转为运动后留下的“鬼影”及光线突变导致的大面积误检情况.采用阴影抑制和形态学滤波使得前景目标分割的性能得到有效的提高.实验表明,本算法能够迅速响应实际场景的变化,准确分割出运动对象.
Abstract:
Aimed at the moving objects segmentation, an algorithm based on improved Gaussian mixture model (GMM) is presented in this paper. The algorithm detects foreground by GMM and classifies it into real changed area and false alarm area. The model sets the update rate in false alarm area quicker than other area to solve "ghost" and sudden illumination change problems due to sudden moving of a motionless object. The performance can be effectively improved with morphology filtering and shadow removal. Experimental results indicate that this algorithm can respond to real scenes quickly, and segment moving objects accurately.

参考文献/References:

[1] WREN C R, AZARBAYE J A, DARRELL T P. Realtime tracking of the human body [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1997, 19(7):780-785.
 [2] STAUFFER C, GRIMSON W E L. Adaptive background mixture models for realtime tracking [C]//Computer Vision and Pattern Recognition. Fort Collins, USA, 1999: 246-250.
 [3] STAUFFER C, GRIMSON W E L. Learning patterns of activity using realtime tracking [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22 (8):747-757.
[4] KAEW TRA KUL PONG P, BOWDEN R. An improved adaptive background mixture model for realtime tracking with shadow detection [C]// The 2nd European Workshop on Adranced Videobased Surveillance Systems. Kingston, 2001:1-5.
[5] YANG Shengyan, HSU Chiouting. Background modeling from GMM likelihood combined with spatial and color coherency \[C\]//2006 IEEE International Conference on ICIP. Atlanta,USA, 2006:2801-2803.
[6] 姚会,苏松志,王丽,等.基于改进的混合高斯模型的运动目标检测方法[J].厦门大学学报,2008, 47(4):505-510.
[7] CHEN Baisheng,LEI Yunqi. Indoor and outdoor people detection and shadow suppression by exploiting HSV color information [C]//Proceedings of the Fourth International Conference on Computer and Information Technology. Wuhan, China, 2004:137-142.
[8] 王勇,谭毅华,金田文.基于阴影消除和混合高斯模型的视频分割算法[J].光电工程,2008, 35(3):21-25.
[9] 刘鑫,刘辉,强振平,等.混合高斯模型和帧间差分相融合的自适应背景模型\[J\].中国图像图形学报,2008, 13(4):729-734.

相似文献/References:

[1]王祖龙,谢红.视频图像火焰区域分割算法[J].应用科技,2011,38(07):16.[doi:doi:10.3969/j.issn.1009-671X.2011.07.04]
 WANG Zulong,XIE Hong.Flame segmentation algorithm based on video image[J].Applied science and technology,2011,38(05):16.[doi:doi:10.3969/j.issn.1009-671X.2011.07.04]
[2]乔玉龙,王玉斐,李娜.基于隐马尔可夫模型的锂电池退化状态识别[J].应用科技,2018,45(02):29.[doi:10.11991/yykj.201612012]
 QIAO Yulong,WANG Yufei,LI Na.Recognition on regression state of lithium-ion battery by using hidden Markov model[J].Applied science and technology,2018,45(05):29.[doi:10.11991/yykj.201612012]
[3]蔡志兆,吕建勋.基于Kinect的运动检测平台设计与实现[J].应用科技,2019,46(03):64.[doi:10.11991/yykj.201807011]
 CAI Zhizhao,LYU Jianxun.Design and implementation of motion detection platform based on Kinect[J].Applied science and technology,2019,46(05):64.[doi:10.11991/yykj.201807011]

备注/Memo

备注/Memo:
作者简介:张宗彬(1985-),男,硕士研究生,主要研究方向:图像处理,E-mail:zhangzongbin@hrbeu.edu.cn.
更新日期/Last Update: 2010-05-26