Parallelization Methods of the Template Matching Method on Graphics Accelerators

被引：0

作者：

Kertesz, Gabor ^{[1
]}

Szenasi, Sandor ^{[1
]}

Vamossy, Zoltan ^{[1
]}

机构：

[1] Obuda Univ, John von Neumann Fac Informat, Budapest, Hungary

来源：

2015 16TH IEEE INTERNATIONAL SYMPOSIUM ON COMPUTATIONAL INTELLIGENCE AND INFORMATICS (CINTI) | 2015年

关键词：

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Template matching is a classic technique used in image processing for object detection. It is based on multiple matrix-based calculations, where there are no dependencies on partial results, so parallel solutions could be created. In this article two GPU implemented methods are presented and compared to the CPU-based sequential solution.

引用

页码：161 / 164

页数：4

共 13 条

[1] Abdel-Hakim A. E., 2005, INF FUS 2005 8 INT C, V2
[2] Almási D, 2014, ACTA POLYTECH HUNG, V11, P169
[3] [Anonymous], 2007, Optimizing parallel reductions in CUDA
[4] [Anonymous], 2016, Programming massively parallel processors: a hands-on approach
[5] [Anonymous], 2014, CUDA C PROGRAMMING G
[6] Brunelli R., 2009, Template Matching Techniques in Computer Vision: Theory and Practice
[7] GROUP AVERAGED LINEAR TRANSFORMS THAT DETECT CORNERS AND EDGES
DUNN, JC
[J]. IEEE TRANSACTIONS ON COMPUTERS, 1975, 24 (12) : 1191 - 1201
[8] Györök G, 2014, ACTA POLYTECH HUNG, V11, P235
[9] Hashimoto T, 2013, ACTA POLYTECH HUNG, V10, P139
[10] I. Corporation and Itseez, 2015, OP 3 0 ONL DOC

← 1 2 →