SIFT for Cirebon Mask Classification

Scale Invariant Feature Transform for Cirebon Mask Classification

Abstract

The vast diversity of art in Indonesia generates much interest both domestically and internationally. One of the prominent cultures is Cirebon Mask. There are five types of Cirebon masks: Panji, Samba, Rumyang, Tumenggung, and Klana. In this research, Cirebon masks are classified using digital images processing techniques using Scale-Invariant Feature Transform, while K-Nearest Neighbour, Support Vector Machines, and Random Forest as the classifier. The best result is obtained by Support Vector Machine and Random Forest.

Methodology

• Data Collection
  • The dataset was collected using two different ways: captured the Cirebon mask images using camera and collected Cirebon mask images using search engine application. This is done to explore more diversity in each class.
• Pre-Processing
  • The background images were removed manually, this makes the background become black. After removing the background images, each image gets resize into 50 x 50 resolution.
• Feature Extraction
  • The SIFT algorithm was implemented to detect key point features from the Cirebon mask. The SIFT algorithm introduced by Lowe. This method is invariant to image translation, partially invariant to affine projection and illumination change, scaling, and rotation. There are four steps for defining SIFT images:
    • step 1: scale-space construction
    • step 2: keypoint localization
    • step 3: orientation
    • step 4: keypoint descriptor
• Classification
  • Feature classification was implemented using K-NN, Random Forest and SVM as the classifier. This classifier created a model for the Cirebon Mask classification.