Temu Kembali Kemiripan Motif Citra Tenun Menggunakan Transformasi Wavelet Diskrit Dan GLCM

Indonesia is a country with cultural diversity. One of the famous cultural heritages in Indonesia is Woven Fabrics. East Nusa Tenggara Province, especially South Central Timor, is an area that also produces weaving. There are 3 types of woven fabric motifs, namely the Buna, Lotis, and Futus motifs w...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Anderias Bai Seran, Aviv Yuniar Rahman, Istiadi Istiadi
Formato: article
Lenguaje:ID
Publicado: Ikatan Ahli Indormatika Indonesia 2021
Materias:
Acceso en línea:https://doaj.org/article/81fe3ec7f87d46e1bd5725c385a8a7a5
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
Descripción
Sumario:Indonesia is a country with cultural diversity. One of the famous cultural heritages in Indonesia is Woven Fabrics. East Nusa Tenggara Province, especially South Central Timor, is an area that also produces weaving. There are 3 types of woven fabric motifs, namely the Buna, Lotis, and Futus motifs which were inherited from their ancestors. Woven cloth is unique because it is made through a ritual process and is used for traditional ceremonies, weddings, funerals, and so on. However, along with the development of technology, ordinary people increasingly forget the motifs of woven fabrics and have difficulty distinguishing the motifs. The function of this research is to improve the performance of previous studies in the process of finding the similarity of weaving image motifs using discrete wavelet transforms and GLCM. The results are known, calculations using a confusion matrix on discrete wavelet transformation feature extraction and GLCM, comparisons on discrete wavelet transformations produce an accuracy rate of 70% Minkowski matrix, 60% Manhattan matrix, 60% Canberra matrix, 20% Euclidean matrix. Comparison of feature extraction calculations on GLCM produces an average quality of the Minkowski matrix of 90% and the lowest level of accuracy on the Euclidean, Manhattan, and Canberra matrices of 80%.