Klasifikasi pada citra Jenis songket Kabupaten Siak Menggunakan Metode Gray Level CO-Occurrence Matrix (GLCM)

Fahrozi, Imam Surya (2021) Klasifikasi pada citra Jenis songket Kabupaten Siak Menggunakan Metode Gray Level CO-Occurrence Matrix (GLCM). Other thesis, Universitas Islam Riau.

Text 163510528.pdf - Submitted Version Download (2MB)

Abstract

Songket is a characteristic of several ethnic groups in Indonesia, songket has many different types and motifs, besides having historical value, the original songket of Siak Regency, namely Songket Pucuk Rebung, Siku Awan, Siku Keluang, and Siku Tampuk Mangosteen, has a level of complexity. which is high in its manufacture, therefore for the identification of songket cloth in Siak Regency in order to make it easier for the public and foreign tourists who come to know the names of songket. An image classification application for the songket type of Siak Regency was developed using the Gray Level Co-Occurrence Matrix (GLCM) method where glcm is used for texture analysis/feature extraction, then classified using the k-Nearest Neighbor (k-NN) Algorithm. Based on the evaluation results of the Songket Pucuk Rebung motif, the class accuracy is 85%, precision is 89%, and recall is 94%. Next is the Siku Cloud Songket motif with 80% accuracy, 84% precision, and 84% recall, the next is the Siku Keluang Songket motif with 75% accuracy, 88% precision, and 0.83% recall, and finally the Siku Tampuk Songket motif. Mangosteen with an accuracy value of 90%, precision of 94%, and recall of 94%.

Actions (login required)

View Item