Spices Image Classification Using Support Vector Machine

Abstract

Spices are ingredients that are commonly used as food flavoring in various types of food in the world. There are many types of spices that exist. Not everyone can distinguish the types of these spices. A Spice classifier is needed to address this problem. This study performs the classification of spice images using the Support Vector Machine (SVM) algorithm, which is a method of mapping data into certain categories. There are 5 spices identified, namely ginger, cinnamon, candlenut, turmeric and pepper. In this study, the SVM model for image identification produced an average accuracy of 38.7%. This model has the greatest accuracy in identifying cinnamon by 65.3% and the smallest accuracy in identifying ginger by 24.5%.