杨大为,刘占林,王琰.大矩阵压缩特征目标的低秩跟踪算法[J].电子测量与仪器学报,2017,31(6):833-838
大矩阵压缩特征目标的低秩跟踪算法
Low rank object tracking algorithm based on large matrix and compressed feature
  
DOI:10.13382/j.jemi.2017.06.003
中文关键词:  压缩感知  矩阵低秩稀疏分解  稀疏表达  增广拉格朗日乘子法  向量相似度
英文关键词:compressed sensing  sparse and low rank matrix decomposition  sparse representation  augmented Lagrange multiplier  vector similarity
基金项目:辽宁省自然科学基金指导计划(2016010993 301)、辽宁省教育厅(LG201609)、沈阳理工大学博士启动(BS 2015 03)资助项目
作者单位
杨大为 沈阳理工大学信息科学与工程学院沈阳110159 
刘占林 沈阳理工大学信息科学与工程学院沈阳110159 
王琰 沈阳理工大学信息科学与工程学院沈阳110159 
AuthorInstitution
Yang Dawei School of Information Science and Engineering, Shenyang Ligong University, Shenyang 110159, China 
Liu Zhanlin School of Information Science and Engineering, Shenyang Ligong University, Shenyang 110159, China 
Wang Yan School of Information Science and Engineering, Shenyang Ligong University, Shenyang 110159, China 
摘要点击次数: 2480
全文下载次数: 15275
中文摘要:
      针对压缩感知的矩阵低秩稀疏分解目标跟踪算法实时性差的问题,提出一种大矩阵压缩特征目标的低秩跟踪算法。该算法通过将大矩阵分成多个小矩阵的方法构建观测矩阵,进行矩阵低秩稀疏分解,获得各候选目标的误差向量并构建误差矩阵,求解误差矩阵列向量最小1 范数问题得到跟踪结果。为了适应跟踪过程中目标外观信息的变化,基于向量相似度判别有选择地更新字典。在跟踪结果不可信时,利用轨迹修正更新当前帧跟踪结果。通过6个典型视频序列上的对比实验,新算法的实时性是原算法的3倍。实验结果表明,在目标发生部分遮挡、光照变化、快速运动时,所提出的算法能实现目标的鲁棒跟踪。
英文摘要:
      In order to improve the instantaneity of the tracking method based on sparse and low rank matrix decomposition, a low rank object tracking algorithm based on large matrix and compressed feature is proposed in this paper. The matrix is decomposed sparsely and low rankly by creating observation matrix using segmenting the large matrix into some parts. Then the error vector of each candidate object is obtained and an error matrix is built. The tracking result is gained by resolving the least 1 norm of the error matrix. To adapt to the changes of target appearance, the dictionary is selectively updated based on the discrimination of vector similarity. When the tracking result is not trusted, it is updated by trajectory rectification. The instantaneity of the new algorithm is three times the old one via the comparison results on six typical sequences. The experiments demonstrate that the proposed algorithm can track the object accurately and robustly when there is part occlusion, illumination change and fast motion.
查看全文  查看/发表评论  下载PDF阅读器