Open AccessDeteksi Batik Parang Menggunakan Fitur Co-Occurence Matrix Dan Geometric Moment Invariant Dengan Klasifikasi KNN
Author(s) -
Ni Luh Wiwik Sri Rahayu Ginantra
Publication year - 2016
Publication title -
lontar komputer/lontar komputer
Language(s) - English
Resource type - Journals
eISSN - 2541-5832
pISSN - 2088-1541
DOI - 10.24843/lkjiti.2016.v07.i01.p05
Subject(s) - artificial intelligence , pattern recognition (psychology) , invariant (physics) , computer science , co occurrence matrix , computer vision , mathematics , image (mathematics) , image texture , image processing , mathematical physics
Batik motifs are the base or the blueprint of batik patterns which serve as the core of the batik image design, and therefore the meaning of a sign, symbol or logo in a batik work can be revealed through its motifs. Visual identification requires visual skills and knowledge in classifying patterns formed in a batik image. Lack of media providing information on batik motifs makes the public unable to have sufficient information about batik motifs. Looking at this phenomenon, this study is conducted in order to perform visual identification using a computer that can assist and facilitate in identifying the types of batik. The methods used for batik image recognition are the Co-occurrence Matrix method to provide extraction of batik texture features, and the Geometric Moment Invariant method, while K Nearest Neighbor is used to classify batik images. The results on the accuracy values obtained reveal that the of 80%, compared to the accuracy value result using the Co-occurrence Matrix method that is 70%.