计算机应用研究2016，Vol.33Issue(8)：2500-2503,4.DOI:10.3969/j.issn.1001-3695.2016.08.058
混合线性模型在模板跟踪中的应用
Mixture hyperplanes approximation for template tracking
摘要
Abstract
Template tracking has been extensively studied in computer vision with a wide range of applications.The motion model is constructed by homography,and the object is tracked by approximating the motion model based on the relationship of the observed data and the motion parameters.This paper proposed a method based on finite mixtures of generalized linear re-gression models to perform robust,real-time tracking from a stationary camera.The model was learnt by motion parameters and appearance data of the object.Moreover,it discussed a fast learning strategy as well,which would improve robustness against noise.It also demonstrated and evaluated the performance and the stability of mixture hyperplanes approximation on a set of challenging image sequences.Experimental result shows that the algorithm has better tracking effect.关键词
模板跟踪/混合线性模型/快速学习Key words
template tracking/mixture hyperplanes approximation/fast learning分类
信息技术与安全科学引用本文复制引用
顾菘,马争,解梅..混合线性模型在模板跟踪中的应用[J].计算机应用研究,2016,33(8):2500-2503,4.基金项目
国家自然科学基金资助项目（61271288）；国家高技术研究发展计划资助项目 （）
