No Detail Found!

Analysis of the effect of feature selection on the handwriting authenticity checking system through fisher score and learning vector quantization

Author(s)
DIAN PRATIWI Department of Informatics Engineering, Universitas Trisakti, Indonesia
SYAFUDIN Department of Informatics Engineering, Universitas Trisakti, Indonesia
TRUBUS RAHARDIANSYAH Department of Law. Universitas Trisakti, Indonesia
MUHAMMAD ICHSAN Department of Informatics Engineering, Universitas Trisakti, Indonesia
STEVEN SEN Department of Informatics Engineering, Universitas Trisakti, Indonesia
AHMAD FAUZY Department of Informatics Engineering, Universitas Trisakti, Indonesia
Abstract
Handwriting is part of the authentic form that everyone has. The authenticity of the owner is often difficult to prove because it is easier to fake with the support of technology. This research was then carried out to develop a method to help verify the authenticity of the owner's handwriting, by applying the LVQ method and selecting the Fisher Score feature to analyze the use of the best number and types of features. The steps taken include processing the grayscale color and data size into 500x200 pixels, then segmentation, taking the features of the writing area based on the mean, variance, entropy, energy and contrast values. The feature is then analyzed by Fisher Score for each use of the feature with the highest score, and tested with LVQ to predict the owner of the article. From the test results, it is found that the use of a large number of features in the order of Fisher's highest score, the quality of accuracy is better than the use of a small number of features
Publication Details
Page(s) 6029-6035
DOI DOI not available
Published Journal: Journal of Theoretical and Applied Information Technology, Volume: 100, Issue: 20, Year: 2022

Request the full text from PASTIC

Keywords
Indexing, Energy, Scoring, Mean, Contrast
0

Citations

0

Downloads

44

Views