Identification of Toga Plants Based on Leaf Image Using the Invariant Moment and Edge Detection Features

Closed

Rosa Andrie Asmara, Mustika Mentari, Nadia Salsabila Herawati Putri, Anik Nur Handayani

2020 4th International Conference on Vocational Education and Training, ICOVET 2020 Conference paper Cited by 6 Quartile

Abstract

Currently, the price of drugs are rising time to time and still growing. The high cost for medical treatment is not compatible compared to the people's welfare, which is burdens to lives of some people. Indonesian people always look for medicine as the main shortcut, whereas in Indonesia, there are a lot of family medicinal garden (in bahasa-TOGA [Tanaman Obat Keluarga]) which are commonly use as herbs and natural medicine. Most people have a difficulty to identify the type of toga plants and the real efficacy of these plants. This research proposed an identification of toga plants using leaf images.The leaf images features will be extract using Invariant Moment and Canny edge will be used to recognize leaf textures. K-Nearest Neighbor is used for leaf type identification. According to the experiments, this system yields 80% classification accuracy. © 2020 IEEE.

Affiliations

State Polytechnic of Malang, Information Technology Department, Malang, Indonesia; Pt Pegadaian (Persero), Quality Assurance Engineer, Jakarta, Indonesia; State University of Malang, Electrical Engineering Department, Malang, Indonesia