Spatiotemporally scalable matrix recovery for background modeling and moving object detection

Spatiotemporally scalable matrix recovery for background modeling and moving object detection
复制标题

用于背景建模和移动物体检测的时空可扩展矩阵恢复

DOI:
10.1016/j.sigpro.2019.107362
复制
发表时间:
2020-03
期刊:
影响因子:
4.4
通讯作者:
Chunping Hou
Chunping Hou
中科院分区:
工程技术2区
文献类型:
--
作者:
Jingyu Yang;Wen Shi;Huanjing Yue;Kun Li;Jian Ma;Chunping Hou

文献摘要

参考文献

相似文献

由于现实世界应用中复杂的移动行为、相机抖动/运动和巨大的数据量,从视频中分离运动对象和背景是视频分析的一项重要但具有挑战性的任务。为了解决这些问题,本文提出了一个统一的框架,称为时空可伸缩矩阵恢复(SSMR),该框架具有中等的计算和空间复杂度,可扩展到视频的时空分辨率。在该模型中,将低秩近似固有的批模式核范数替换为显式的低秩矩阵分解,以实现在线实现。将光流法提取的运动信息合并到数据项中,便于将运动物体与背景分离。仿射变换嵌入到模型中,并与其他变量同时优化以处理相机运动。此外,我们提出了一种金字塔结构来实现高清晰度视频的空间可扩展性。实验结果表明,我们的方法优于许多其他先进的方法,可以处理各种复杂场景的视频。
Separating moving objects and backgrounds from a video is an important yet challenging task for video analysis due to complex moving behaviors, camera jitters/movements, and huge data amount in real-world applications. To deal with these issues, this paper proposes a unified framework called spatiotemporally scalable matrix recovery (SSMR), which has a moderate computational and space complexity scalable to temporal and spatial resolution of videos. In the proposed model, the inherent batch-mode nuclear norm for low-rank approximation is replaced with an explicitly low-rank matrix factorization in order to achieve online implementation. Motion information extracted by an optical flow method is incorporated into the data term to facilitate the separation of moving objects from the background. Affine transformation is embedded into the model and simultaneously optimized with other variables to handle camera motions. In addition, we proposed a pyramidal scheme to achieve spatial scalability for high definition videos. Experimental results demonstrate that our method outperforms many other state-of-the-art methods and can handle videos of various complex scenarios.
DOI: 10.1109/vcip.2016.7805552
发表时间: 2016-11
期刊: 2016 Visual Communications and Image Processing (VCIP)
影响因子: --
作者:
Jiaoru Yang;Jingyu Yang;Xuemeng Yang;Huanjing Yue
通讯作者: Jiaoru Yang;Jingyu Yang;Xuemeng Yang;Huanjing Yue
DOI: 10.1109/tsp.2003.815389
发表时间: 2003-09-01
影响因子: 5.4
作者:
Do, MN;Vetterli, M
通讯作者: Vetterli, M
DOI: 10.1109/34.598236
发表时间: 1997-07-01
影响因子: 23.6
作者:
Wren, CR;Azarbayejani, A;Pentland, AP
通讯作者: Pentland, AP
DOI: 10.1109/tcsvt.2017.2733623
发表时间: 2018-11
影响因子: 8.4
作者:
Hong-Cyuan Wang;Yu-Chi Lai;Wen-Huang Cheng;C. Cheng;K. Hua
通讯作者: Hong-Cyuan Wang;Yu-Chi Lai;Wen-Huang Cheng;C. Cheng;K. Hua
DOI: 10.1109/afgr.1996.557243
发表时间: 1996-01
期刊: Proceedings of the Second International Conference on Automatic Face and Gesture Recognition
影响因子: --
作者:
C. Wren;A. Azarbayejani;Trevor Darrell;A. Pentland
通讯作者: C. Wren;A. Azarbayejani;Trevor Darrell;A. Pentland