IMPLEMENTASI DENSITY-BASED CLUSTERING PADA SEGMENTASI CITRA Betta Fish

Abstract

During the COVID-19 pandemic, the number of ornamental fish enthusiasts has increased, especially those of the Betta Fish species. Betta Fish is a type of ornamental Fish with various species with beautiful colors and morphology, especially the shape of the tail. The more diverse the color patterns of the Fish and the unique shape, the more expensive the selling price of this type of ornamental Fish. The market demand for Betta Fish is getting higher, causing the selling price of Betta Fish also to increase. However, not all ornamental fish lovers recognize the species name of the Betta Fish. For this reason, a pattern recognition-based system is needed that can recognize Betta Fish species. Pattern recognition has several stages, namely segmentation, extraction, and classification. This study aims to separate the object from the background on a digital image. The dataset used is 160 images consisting of 40 images of each species, namely Halfmoon, Double Tail, Serit, and Plakat. The first step is to convert the image into a saturation and intensity color model. The method used in the segmentation process is Density-Based Clustering. Density-Based Clustering is a segmentation method by forming clusters based on the density level of the object area. The segmentation process using the DensityBased Clustering method achieves an accuracy rate of 92.82%.