Klasifikasi Citra Makanan Menggunakan K-Nearest Neighbor dengan Fitur Bentuk Simple Morphological Shape Descriptors dan Fitur Warna Grayscale Histogram

Muhammad Rizky Setiawan (Fakultas Ilmu Komputer, Universitas Brawijaya)
Yuita Arum Sari (Fakultas Ilmu Komputer, Universitas Brawijaya)
Putra Pandu Adikara (Fakultas Ilmu Komputer, Universitas Brawijaya)

Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Vol 3 No 3 (2019): Maret 2019

Food is one of the energy sources needed by humans. The type of food consumed greatly affect the immune system. But the diversity of existing food causes people to be difficult to recognize the type of food they want to consume. The need for a system that can recognize types of food to make it easier for people to regulate their diet. Before entering the feature extraction process, the first step is to do preprocessing by separating the background from the food image object. Furthermore, color feature extraction is performed using the Grayscale Histogram method. The Grayscale Histogram method produces the mean, standard deviation, skewness features. Then form feature extraction was performed using the Simple Morphological Shape Descriptors (SMSD) method and produced area features, length, width, aspect ratio, rectangular N. After extracting feature results, classification was done using the K-Nearest Neighbor method. Based on the test results if only using the Grayscale Histogram method produces an accuracy value of 60%. If only using the SMSD method produces an accuracy value of 54.8%. If using the Grayscale Histogram method and the SMSD method produces an accuracy value of 77.8%. The Grayscale Histogram method and the SMSD method can be used to process images using the K-Nearest Neighbor classification method.