《计算机应用研究》|Application Research of Computers

基于部分背景加权更新的均值漂移跟踪算法

Mean-Shift tracking algorithm based on background weighted update approach

免费全文下载 (已被下载 次)  
获取PDF全文
作者 李茂林,刘小平,胡凌燕,徐少平
机构 1.南昌大学 信息工程学院,南昌 330031;2.加拿大卡尔顿大学 系统与计算机工程系,渥太华 K1S5B6
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2013)12-3810-04
DOI 10.3969/j.issn.1001-3695.2013.12.076
摘要 为了解决均值漂移跟踪算法中背景对目标定位的扰动, 提出了一种基于颜色和纹理混合特征以及采用背景加权更新的改进算法。改进算法先将原始视频序列RGB帧图像转换为HSV颜色空间表示, 然后分别在H、S通道上提取颜色特征, 在V通道上用LBP描述符提取纹理特征, 在此基础上为目标区域和背景区域建立三维颜色纹理混合直方图作为其描述符; 在对象的跟踪过程中, 通过巴氏系数选择性地加权更新部分背景信息。实验结果表明, 与基于全部背景更新策略相比, 改进算法充分利用了颜色和纹理特征并加权更新背景信息, 具有更高的可靠性和鲁棒性, 具有更好的计算效率。
关键词 目标跟踪;均值漂移;HSV空间;局部二值模式;加权背景更新
基金项目 国家“863”计划资助项目(2013AA013804)
国家自然科学基金资助项目(61175072,61163023,51165033)
江西省科技支撑计划资助项目(20121BBE50023)
本文URL http://www.arocmag.com/article/01-2013-12-076.html
英文标题 Mean-Shift tracking algorithm based on background weighted update approach
作者英文名 LI Mao-lin, LIU Xiao-ping, HU Ling-yan, XU Shao-ping
机构英文名 1. School of Information Engineering, Nanchang University, Nanchang 330031, China; 2. Dept. of Systems & Computer Engineering, Carleton University, Ottawa K1S5B6, Canada
英文摘要 In order to solve the problems of background interference in the Mean-Shift tracking algorithm, this paper proposed an improved algorithm based on color and texture blending characteristics and background weighted update approach. The original RGB image was converted into the HSV color space, then color feature was extracted in the H, S channel and texture feature was extracted based on the LBP descriptor in the V channel. Base on this, this paper established the color-texture histogram of the object region and background. During object tracking, it updated the background region using weighted update approach according to the Bhattacharyya coefficient. The extensive experimental results show that, compared with the algorithm adopting the full background update approach with color or color-texture features, the improved algorithm makes full use of color and texture features and adopts weighted updated background region, and has a higher level of reliability and robustness and better execution efficiency.
英文关键词 target location; Mean-Shift; HSV space; local binary pattern; weighted background update
参考文献 查看稿件参考文献
  [1] COMANICIU D, RAMESH V, MEER P. Kernel-based object trac-king[J] . IEEE Trans on Pattern Analysis and Machine Intelligence, 2003, 25(5):564-577.
[2] LEICHTER I, LINDENBAUM M, RIVLIN E. Mean shift tracking with multiple reference color histograms[J] . Image and Vision Computing, 2009, 27(5):535-544.
[3] 郑玉凤, 马秀荣, 赵晓琳, 等. 基于颜色和边缘特征的均值迁移目标跟踪算法[J] . 光电子·激光, 2011, 22(8):1231-1235.
[4] 戴渊明, 韦巍, 林亦宁. 基于颜色纹理特征的均值漂移目标跟踪算法[J] . 浙江大学学报:工学版, 2012, 46(2):212-217.
[5] NING Ji-feng, ZHANG Lei, ZHANG D, et al. Robust object tracking using joint color-texture histogram[J] . International Journal of Pattern Recognition and Artificial Intelligence, 2009, 23(9):1245-1263.
[6] HEIKKIL M, PIETIKINEN M, SCHMID C. Description of interest regions with local binary patterns[J] . Pattern Recognition, 2009, 42(3):425-426.
[7] TAN X, TRIGGS B. Enhanced local texture feature sets for face recognition under difficult lighting conditons[J] . Image Processing, 2010, 19(6):1635-1650.
[8] 袁国武, 徐丹. 一种结合了纹理和颜色的运动目标跟踪算法[J] . 计算机应用与软件, 2011, 28(11):81-84.
[9] 宁纪锋, 吴成柯. 一种基于纹理模型的目标跟踪算法[J] . 模式识别与人工智能, 2007, 20(5):612-618.
[10] NING Ji-feng, ZHANG Lei, ZHANG D, et al. Robust Mean-Shift tracking with corrected background-weighted histogram[J] . Computer Vision, 2012, 6(1):62 -69.
[11] OJALA T, PIETIKINEN M, MENP T. Multiresolution gray-scale and rotate invariant texture classification with local binary patterns[J] . Pattern Analysis and Machine Intelligence, 2002, 24(7):971-987.
[12] OJALA T, VALKEALAHTI K, OJA E, et al. Texture discrimination with multidimensional distributions of signed gray level differences[J] . Pattern Recognition, 2001, 34(3):727-739.
[13] 李冠彬, 吴贺丰. 基于颜色纹理直方图的带权分块均值漂移目标跟踪算法[J] . 计算机辅助设计与图形学学报, 2011, 23(12):2059-2066.
[14] WANG Ling-feng, PAN Chun-hong, XIANG Shi-ming. Mean-Shift tracking algorithm with weight fusion strategy[C] //Proc of the 18th International Conference on Image Processing. 2011:473-476.
[15] VILAPLANA V, VARAS D. Face tracking using a regionbased Mean-Shift algorithm with adaptive object and background models[C] //Proc of the 10th International Workshop on Image Analysis for Multimedia Interactive Services. Washington DC:IEEE Computer Society, 2009:9-12.
收稿日期
修回日期
页码 3810-3813
中图分类号 TP391;TP301.6
文献标志码 A