A New Object Tracking Algorithm Based on the Fast Discrete Curvelet Transform
A New Object Tracking Algorithm Based on the Fast Discrete Curvelet Transform
复制标题
一种基于快速离散曲波变换的目标跟踪新算法
DOI：
10.14257/ijsip.2014.7.1.06
复制
发表时间：
2014
期刊：
影响因子：
--
通讯作者：
Liying Zheng
中科院分区：
文献类型：
--
作者：
Lei Yu;Xin;Liying Zheng
从视频序列中跟踪目标已经成为计算机视觉领域中一个非常热门的问题。已经提出了许多算法来解决这个问题。结合曲线系数能量和卡尔曼滤波方法，提出了一种基于快速离散曲线变换的目标跟踪算法。首先，通过建立运动目标模板对算法进行初始化，并利用搜索策略得到若干未确定目标；然后，利用曲线系数的能量构造特征向量；最后，利用相似度函数匹配特征向量，在下一帧中跟踪运动目标。实验结果证明了该算法在遮挡下的跟踪精度和效率。
Tracking object from video sequences has become a very popular problem in the field of computer vision. Many algorithms have been proposed to solve this problem. Combined with the energy of curvelet coefficients and Kalman filter method, a new object tracking algorithm is proposed based on fast discrete curvelet transform. Firstly, this algorithm is initialized with the establishment of a moving object template and a number of undetermined objects are got with the search strategy. Then, feature vectors are constructed by energy of curvelet coefficients. Finally, the moving object is tracked in next frames using similarity function to match the feature vectors. Experimental results demonstrate the tracking accuracy and efficiency of the proposed algorithm under occlusions.
DOI：
10.1109/34.1000236
发表时间：
2002-05-01
影响因子：
23.6
作者：
Comaniciu, D;Meer, P
通讯作者：
Meer, P