Open AccessConsumable Fish Classification Using k-Nearest Neighbor
Author(s) -
Sri Winiarti,
Fitri Indra Indikawati,
Ardyawati Wira Oktaviana,
Herman Yuliansyah
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/821/1/012039
Subject(s) - mackerel , hue , artificial intelligence , pattern recognition (psychology) , fish products , tilapia , classifier (uml) , grayscale , fish <actinopterygii> , computer science , fishery , mathematics , biology , image (mathematics)
Fish is beneficial for the human body because it has high protein content. Consuming fish is necessary and expert knowledge is needed to identify fresh fish that are suitable for consumption. In this study, we developed a classification system to identify four classes of consumable fish by grouping fish images based on texture extraction and color features. We use fish meat and fish scale as identification parameters. Fish meat image is measured using the HSV colors model (Hue, Saturation, and Value) and GLCM (Gray Level Co-occurrence Matrix) method. We use these values for texture feature extraction of scales. Then we use k-Nearest Neighbor (kNN) as the classifier. The test results from 320 sample images show that the identification accuracy of tilapia meat is 90% and 97.5% for mackerel meat. Meanwhile for the scales, the accuracy up to 87.5% for tilapia scales and 95% for mackerel scales.