SMARTFLORA Mobile Flower Recognition Application Using Machine Learning Tools

Abstract

There are around 369,000 flowering plant species documented globally. However, the majority of people have difficulties telling these blooms apart. Usually, people often consult specialists, study floral reference books, or do keyword searches on relevant web resources. Therefore, this flower recognition mobile application was proposed to ease those people to recognize types of flowers without using any computer or machine. In this paper, a system architecture is designed based on Teachable Machine Learning platform, Tensorflow Lite Model and Android Studio to develop a SMARTFLORA Mobile Flower Recognition application that allows users to identify three types of flower species: daisies, roses, and sunflowers. Kaggle dataset has been used and the accuracy was 88%.