[1]Zhang Hongmei,Wen Hueran,Zhang Xiangli,et al.Sparse representation tracking via compressed features[J].Journal of Zhengzhou University (Engineering Science),2016,37(03):21-26.[doi:10.13705/j.issn.1671-6833.2016.03.005]
Copy
Journal of Zhengzhou University (Engineering Science)[ISSN
1671-6833/CN
41-1339/T] Volume:
37
Number of periods:
2016 03
Page number:
21-26
Column:
Public date:
2016-05-10
- Title:
-
Sparse representation tracking via compressed features
- Author(s):
-
Zhang Hongmei; Wen Hueran; Zhang Xiangli; Li Pengfei
-
School of Guilin University of Electronic Technology, Guilin, Guangxi, 541004
-
- Keywords:
-
features compression; sparse representation; particle filter; block orthogonal matching
- CLC:
-
TP391
- DOI:
-
10.13705/j.issn.1671-6833.2016.03.005
- Abstract:
-
In order to deal with the influence of the factors such as light, shade, movement of object and etc.,the integral graph method is used to extract the Haar-like features of the target template, and the features arecompressed by a random sparse matrix which meets the limited equidistant conditions( RIP), then the con-struction of the target features dictionary is simplified. Meanwhile, the background information is added in thedictionary, and the simple relationship between the target and the background is used to improve the accuracyof tracking. At last, the target can be reconstructed in block by using the block orthogonal matching pursuit(BOMP)reconstruction algorithm, through which can enhances tracking speed. The experimental results showthat, the block orthogonal matching pursuit tracking algorithm based on compression feature is powerful in val-id target appearance model construction. And it also enhances the tracking stability and improves trackingspeed.