结合特征在线选择与协方差矩阵的压缩跟踪算法  

Compressive tracking algorithm combining online feature selection with covariance matrix

作　　者：张红颖[1] 李灿锋[1] 

机构地区：[1]中国民航大学电子信息与自动化学院,天津300300

出　　处：《光学精密工程》2017年第4期1051-1059,共9页Optics and Precision Engineering

基　　金：天津市自然科学基金青年基金资助项目(12JCQNJC00600);中央高校基本科研业务费(3122015C016)

摘　　要：本文从3个方面对原始压缩跟踪算法进行改进,以提高其在复杂场景下的鲁棒性和准确性。首先,提出一种结合特征在线选择的压缩跟踪算法,通过计算相邻两帧同维特征所服从的高斯分布曲线的Hellinger距离来度量特征的置信水平,从特征池中选择置信水平较高的特征,并融合特征的置信水平构造贝叶斯分类器。然后,在压缩跟踪框架下引入协方差矩阵以增强算法对目标的表达能力,把Haar-like特征和协方差矩阵相结合构建目标模型,取最大响应值所对应的候选样本作为跟踪结果。最后,优化分类器参数的更新方式,根据目标模板与跟踪结果的相似度来自适应更新分类器参数。改进算法的平均跟踪成功率比原算法提高了25%,平均跟踪精度比原算法提高了22%。相比于原始压缩跟踪算法,本文算法具有更高的跟踪鲁棒性和准确性。To improve robustness and accuracy of original compressive tracking algorithm in complex scenes, improvement measures were carried out from three aspects in this paper. First, a compressive tracking algorithm combining with online feature selection was introduced, then the confidence level of the feature was measured by calculating Hellinger distance in the Gaussian distribution curve to which the same dimensional features of two adjacent frames. By selecting the feature with higher confidence level from the feature pool to construct the Bayesian classifier through integrating the confidence lev- el; then, covariance matrix was introduced under the compressive tracking framework to enhance ex- pressive ability of the algorithm towards the target, subsequently, combined Haar--like with covari- ance matrix to create the target model and selected the candidate sample corresponding to the maxi- mum response value as the tracking results; finally, updated mode of the classifier parameters was op- timized： adaptively updating of the classifier parameters was implemented in accordance with similari-ty between the target template and tracking results. It indicates that compared with original algo- rithm, average success rate of proposed algorithm is improved by 25~//oo, and the average tracking accu- racy is improved by 22~. Hence, the algorithm proposed in this paper can achieve higher robustness and accuracy than the original compressive tracking algorithm.

关 键 词：压缩跟踪 特征在线选择 协方差矩阵 分类器 自适应更新 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]