A Parallel Optimization Method for KCF based on Inter-core Communication of Multi-core DSP

Wancang Wu; Qingzhong Jia; Fengkun Luo; Tao Yang; Hong Huang

doi:10.1109/IMCEC51613.2021.9482038

A Parallel Optimization Method for KCF based on Inter-core Communication of Multi-core DSP

Wancang Wu^*, Qingzhong Jia, Fengkun Luo, Tao Yang, Hong Huang

^*Corresponding author for this work

School of Aerospace Engineering

Beijing Institute of Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Target tracking is an important branch in the field of computer vision. In recent years, as target tracking application have increasingly higher requirements for realtime performance, higher requirements have been put forward for the optimization and application of embedded platforms. Therefore, in a multi-core DSP platform, parallel computing efficiency becomes the key to improving realtime performance. This paper proposes the concept of execution parallelism based on inter-core communication, and optimizes KCF for the parallelism parameters. This paper aims at the multi-core parallel development of KCF algorithm based on inter-core interrupt and shared memory on TMS320C6678 platform. By connecting the FPGA-based video board for physical verification, the parallel scheme based on inter-core communication can reach a tracking speed of 60 fps, which has certain engineering practical significance.

Original language	English
Title of host publication	IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference
Editors	Bing Xu
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	1736-1740
Number of pages	5
ISBN (Electronic)	9781728185347
DOIs	https://doi.org/10.1109/IMCEC51613.2021.9482038
Publication status	Published - 18 Jun 2021
Event	4th IEEE Advanced Information Management, Communicates, Electronic and Automation Control Conference, IMCEC 2021 - Chongqing, China Duration: 18 Jun 2021 → 20 Jun 2021

Publication series

Name	IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference
ISSN (Print)	2693-2814
ISSN (Electronic)	2693-2776

Conference

Conference	4th IEEE Advanced Information Management, Communicates, Electronic and Automation Control Conference, IMCEC 2021
Country/Territory	China
City	Chongqing
Period	18/06/21 → 20/06/21

Keywords

KCF
multi-core DSP
parallel optimization
pipeline parallel architecture

Access to Document

10.1109/IMCEC51613.2021.9482038

Cite this

Wu, W., Jia, Q., Luo, F., Yang, T., & Huang, H. (2021). A Parallel Optimization Method for KCF based on Inter-core Communication of Multi-core DSP. In B. Xu (Ed.), IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference (pp. 1736-1740). (IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IMCEC51613.2021.9482038

Wu, Wancang ; Jia, Qingzhong ; Luo, Fengkun et al. / A Parallel Optimization Method for KCF based on Inter-core Communication of Multi-core DSP. IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference. editor / Bing Xu. Institute of Electrical and Electronics Engineers Inc., 2021. pp. 1736-1740 (IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference).

@inproceedings{9a0b24b394134c0a8a9a5c1602dc445e,

title = "A Parallel Optimization Method for KCF based on Inter-core Communication of Multi-core DSP",

abstract = "Target tracking is an important branch in the field of computer vision. In recent years, as target tracking application have increasingly higher requirements for realtime performance, higher requirements have been put forward for the optimization and application of embedded platforms. Therefore, in a multi-core DSP platform, parallel computing efficiency becomes the key to improving realtime performance. This paper proposes the concept of execution parallelism based on inter-core communication, and optimizes KCF for the parallelism parameters. This paper aims at the multi-core parallel development of KCF algorithm based on inter-core interrupt and shared memory on TMS320C6678 platform. By connecting the FPGA-based video board for physical verification, the parallel scheme based on inter-core communication can reach a tracking speed of 60 fps, which has certain engineering practical significance.",

keywords = "KCF, multi-core DSP, parallel optimization, pipeline parallel architecture",

author = "Wancang Wu and Qingzhong Jia and Fengkun Luo and Tao Yang and Hong Huang",

note = "Publisher Copyright: {\textcopyright} 2021 IEEE.; 4th IEEE Advanced Information Management, Communicates, Electronic and Automation Control Conference, IMCEC 2021 ; Conference date: 18-06-2021 Through 20-06-2021",

year = "2021",

month = jun,

day = "18",

doi = "10.1109/IMCEC51613.2021.9482038",

language = "English",

series = "IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "1736--1740",

editor = "Bing Xu",

booktitle = "IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference",

address = "United States",

}

Wu, W, Jia, Q, Luo, F, Yang, T & Huang, H 2021, A Parallel Optimization Method for KCF based on Inter-core Communication of Multi-core DSP. in B Xu (ed.), IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference. IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference, Institute of Electrical and Electronics Engineers Inc., pp. 1736-1740, 4th IEEE Advanced Information Management, Communicates, Electronic and Automation Control Conference, IMCEC 2021, Chongqing, China, 18/06/21. https://doi.org/10.1109/IMCEC51613.2021.9482038

A Parallel Optimization Method for KCF based on Inter-core Communication of Multi-core DSP. / Wu, Wancang; Jia, Qingzhong; Luo, Fengkun et al.
IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference. ed. / Bing Xu. Institute of Electrical and Electronics Engineers Inc., 2021. p. 1736-1740 (IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - A Parallel Optimization Method for KCF based on Inter-core Communication of Multi-core DSP

AU - Wu, Wancang

AU - Jia, Qingzhong

AU - Luo, Fengkun

AU - Yang, Tao

AU - Huang, Hong

PY - 2021/6/18

Y1 - 2021/6/18

N2 - Target tracking is an important branch in the field of computer vision. In recent years, as target tracking application have increasingly higher requirements for realtime performance, higher requirements have been put forward for the optimization and application of embedded platforms. Therefore, in a multi-core DSP platform, parallel computing efficiency becomes the key to improving realtime performance. This paper proposes the concept of execution parallelism based on inter-core communication, and optimizes KCF for the parallelism parameters. This paper aims at the multi-core parallel development of KCF algorithm based on inter-core interrupt and shared memory on TMS320C6678 platform. By connecting the FPGA-based video board for physical verification, the parallel scheme based on inter-core communication can reach a tracking speed of 60 fps, which has certain engineering practical significance.

AB - Target tracking is an important branch in the field of computer vision. In recent years, as target tracking application have increasingly higher requirements for realtime performance, higher requirements have been put forward for the optimization and application of embedded platforms. Therefore, in a multi-core DSP platform, parallel computing efficiency becomes the key to improving realtime performance. This paper proposes the concept of execution parallelism based on inter-core communication, and optimizes KCF for the parallelism parameters. This paper aims at the multi-core parallel development of KCF algorithm based on inter-core interrupt and shared memory on TMS320C6678 platform. By connecting the FPGA-based video board for physical verification, the parallel scheme based on inter-core communication can reach a tracking speed of 60 fps, which has certain engineering practical significance.

KW - KCF

KW - multi-core DSP

KW - parallel optimization

KW - pipeline parallel architecture

UR - http://www.scopus.com/inward/record.url?scp=85116081564&partnerID=8YFLogxK

U2 - 10.1109/IMCEC51613.2021.9482038

DO - 10.1109/IMCEC51613.2021.9482038

M3 - Conference contribution

AN - SCOPUS:85116081564

T3 - IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference

SP - 1736

EP - 1740

BT - IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference

A2 - Xu, Bing

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 4th IEEE Advanced Information Management, Communicates, Electronic and Automation Control Conference, IMCEC 2021

Y2 - 18 June 2021 through 20 June 2021

ER -

Wu W, Jia Q, Luo F, Yang T, Huang H. A Parallel Optimization Method for KCF based on Inter-core Communication of Multi-core DSP. In Xu B, editor, IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference. Institute of Electrical and Electronics Engineers Inc. 2021. p. 1736-1740. (IMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference). doi: 10.1109/IMCEC51613.2021.9482038

A Parallel Optimization Method for KCF based on Inter-core Communication of Multi-core DSP

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this