Identifikasi Daun Obat-Obatan menggunakan Ssd-Mobilenet secara Realtime

Abstract

Medicinal plants are very easy to grow and cultivated in Indonesia. Many species are even found thriving wild on the roadsides and bushes, but are difficult to identify with certainty. The use of deep learning today has helped many developments in the field of computer vision and object identification. The purpose of this research is to facilitate the identification stage of medicinal plants. The method used is SSD- MobileNet in real time with research objects namely bandotan (Ageratum conyzoides), meniran (Phyllanthus urinaria), and Chinese betel (Peperomia pellucida). This research results showed that SSD-MobileNet was able to identify objects with accuracy rate up to 92.7%.