[1]LIU Jiwei,WEI Honglei,PEI Qichao,et al.Underwater sea cucumber target tracking algorithm based on correlation filtering[J].CAAI Transactions on Intelligent Systems,2019,14(3):525-532.[doi:10.11992/tis.201711037]
Copy
CAAI Transactions on Intelligent Systems[ISSN 1673-4785/CN 23-1538/TP] Volume:
14
Number of periods:
2019 3
Page number:
525-532
Column:
学术论文—智能系统
Public date:
2019-05-05
- Title:
-
Underwater sea cucumber target tracking algorithm based on correlation filtering
- Author(s):
-
LIU Jiwei1; WEI Honglei1; PEI Qichao1; XING Liran2
-
1. Institute of Mechanical Engineering and Automation, Dalian Polytechnic University, Dalian 116034, China;
2. College of Mechanical Engineering, North China University of Science and Technology, Tangshan 063210, China
-
- Keywords:
-
visual tracking; circulant matrices; discrete Fourier transform; kernel methods; ridge regression; correlation filters; capturing sea cucumbers; scale estimation
- CLC:
-
TP391
- DOI:
-
10.11992/tis.201711037
- Abstract:
-
This study proposes a type of sea cucumber target tracking algorithm based on the kernel correlation filter (KCF) to find a solution for real-time tracking while capturing a sea cucumber using an underwater robot. In the initial frame, the image block that contains the target sea cucumber is divided into nine sub-blocks based on the characteristics of the sea cucumber, including its appearance and the positioning of its two heads by comparing the boundary blocks with the central block. Further, the KCF algorithm is used to track the two heads of the sea cucumber in the subsequent frames, estimate the scale, and calculate the location of the sea cucumber based on the distance variation between the two modules. The experimental results exhibit that the accuracy, running speed, and success rate of the tracking algorithm are higher than those of other experimental algorithms.