Rhizome Image Classification Using Support Vector Machine

Abstract

Rhizomes, also called rootstalks, are stems that help plants to reproduce asexually, survive in winter, store food, and make stem tubers.They possess many functions and merits.Some of the commonly found rhizomes are ginger, turmeric, and galangal, yet still a lot of people still find it difficult to distinguish those three rhizomes.That's because those mentioned rhizomes do share several similarities in their shape and texture.This research submits a rhizome identification system with SVM (Support Vector Machine) classification method.Based on the experiments done, this particular method is chosen because it showed great results, quite high-valued accuracy level for data classification, and has minimum error rate.The types of rhizomes used in this research's dataset are those three varieties mentiones above, while the amount of images in this experiment consists of 150 training images and 30 testing images.The experiment is done by calculating the accuracy value from data testing classification of three classes, which are ginger class, turmeric class, and galangal class utilizing the mentioned method.This rhizome identification system that uses the SVM classificafion method gets 78% accuracy value.