Correlation-Coefficient-Based Fast Template Matching Through Partial Elimination

被引:83
作者
Mahmood, Arif [1 ]
Khan, Sohaib [2 ]
机构
[1] Univ Punjab, Coll Informat Technol, Lahore 54000, Pakistan
[2] Lahore Univ Management Sci, Sch Sci & Engn, Lahore 54810, Pakistan
关键词
Image matching; image recognition; image registration; ALGORITHMS;
D O I
10.1109/TIP.2011.2171696
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Partial computation elimination techniques are often used for fast template matching. At a particular search location, computations are prematurely terminated as soon as it is found that this location cannot compete with an already known best match location. Due to the nonmonotonic growth pattern of the correlation-based similarity measures, partial computation elimination techniques have been traditionally considered inapplicable to speed up these measures. In this paper, we show that partial elimination techniques may be applied to a correlation coefficient by using a monotonic formulation, and we propose basic-mode and extended-mode partial correlation elimination algorithms for fast template matching. The basic-mode algorithm is more efficient on small template sizes, whereas the extended mode is faster on medium and larger templates. We also propose a strategy to decide which algorithm to use for a given data set. To achieve a high speedup, elimination algorithms require an initial guess of the peak correlation value. We propose two initialization schemes including a coarse-to-fine scheme for larger templates and a two-stage technique for small- and medium-sized templates. Our proposed algorithms are exact, i.e., having exhaustive equivalent accuracy, and are compared with the existing fast techniques using real image data sets on a wide variety of template sizes. While the actual speedups are data dependent, in most cases, our proposed algorithms have been found to be significantly faster than the other algorithms.
引用
收藏
页码:2099 / 2108
页数:10
相关论文
共 32 条
[1]   CLASS OF ALGORITHMS FOR FAST DIGITAL IMAGE REGISTRATION [J].
BARNEA, DI ;
SILVERMAN, HF .
IEEE TRANSACTIONS ON COMPUTERS, 1972, C 21 (02) :179-+
[2]   AN IMPROVEMENT OF THE MINIMUM DISTORTION ENCODING ALGORITHM FOR VECTOR QUANTIZATION [J].
BEI, CD ;
GRAY, RM .
IEEE TRANSACTIONS ON COMMUNICATIONS, 1985, 33 (10) :1132-1133
[3]   A SURVEY OF IMAGE REGISTRATION TECHNIQUES [J].
BROWN, LG .
COMPUTING SURVEYS, 1992, 24 (04) :325-376
[4]  
CHALERMWAT P, 1999, THESIS G MASON U FAI
[5]   Adjustable partial distortion search algorithm for fast block motion estimation [J].
Cheung, CH ;
Po, LM .
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, 2003, 13 (01) :100-110
[6]   The design and implementation of FFTW3 [J].
Frigo, M ;
Johnson, SG .
PROCEEDINGS OF THE IEEE, 2005, 93 (02) :216-231
[7]  
Frigo M, 1998, INT CONF ACOUST SPEE, P1381, DOI 10.1109/ICASSP.1998.681704
[8]   A fast Fourier transform compiler [J].
Frigo, M .
ACM SIGPLAN NOTICES, 1999, 34 (05) :169-180
[9]  
GOSHTASBY A, 1984, IEEE T PATTERN ANAL, V6, P374, DOI 10.1109/TPAMI.1984.4767532
[10]  
Haralick R. M., 1992, Computer and Robot Vision