
Templates selection in non‐overlapping template matching test
Author(s) -
Chen Meihui,
Chen Hua,
Fan Limin,
Feng Dengguo
Publication year - 2016
Publication title -
electronics letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.375
H-Index - 146
ISSN - 1350-911X
DOI - 10.1049/el.2016.0260
Subject(s) - template , randomness , selection (genetic algorithm) , correlation , template matching , matching (statistics) , mathematics , random number generation , computer science , algorithm , statistics , artificial intelligence , geometry , image (mathematics) , programming language
It is important to study the correlation of randomness statistical tests The correlation of templates in non‐overlapping template matching test is studied. The correlation degree between templates is defined and a correlation theorem is presented. For two different m ‐bit templates, when the last m −1 bits of one are equal to the first m −1 bits of the other one, correlation degree of them reaches the maximum value. Correlation experiments are also carried out and the results agree with the theorem. Finally, an effective selection strategy is proposed and the number of templates can be reduced about 50%.