Open AccessA Signature Best feature selection matching using FAST and genetic algorithm
Author(s) -
Mohammad Q. Jawad,
Tamara Z. Fadhil
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/870/1/012132
Subject(s) - computer science , biometrics , signature (topology) , pattern recognition (psychology) , feature selection , matching (statistics) , feature (linguistics) , artificial intelligence , data mining , feature extraction , process (computing) , artificial neural network , matlab , genetic algorithm , algorithm , machine learning , mathematics , linguistics , statistics , philosophy , geometry , operating system
Verification for human is an important method for identifying persons. One of the most important biometric verification is signature which is used to ensure human privacy, both in banking and electronic business signature is the main authentication step for verify the user identity. In this paper system be tested by genetic algorithm to choose the best feature to applied in training and testing which will reduce the require time needed for matching and training and to obtain better matching by checking only the strong features. The feature selection algorithm allows us to reduce the process time and have more accurate results. The GA diagram was acquainted and applied with an element Issue choosing a subset to approve the mark The calculation demonstrated striking execution in all tests led System split mainly to many stages which is mainly (pre-processing, feature extraction using FAST (Fast Affine Invariant Image Matching), feature reduction (best feature selection) using genetic algorithm and matching process using neural network) The system was tested with 240 signature images with accuracy of 87% with the matching process. The system is simulated and tested using MATLAB.