Plant Recognition and its Care using TfLite, Teachable Machine

Abstract

Abstract: Plant identification is a significant undertaking however its care is considerably more significant. The identification of plants has a lot of benefits for wide range of people ranging from forestry services, pharmaceutical labs, government and common public. The research paper aims on developing a machine learning technology combined with mobile app development which is able to recognize plants and suggest good care for them like how much water, sunlight, fertilizer, etc is optimum for the plant. Plant identification with its care, based on leaf, is becoming an interesting task. Each plant leaf carries unique identity, which matches to no other plant, information that can be used in the identification and subsequently, determining measures for its care. A large number of features were taken from each leaf such as its length, area of hull, perimeter and color. All the work of research in this filed has further aggravated an interest in the development of automated systems for the recognition of different plant species. A full autonomous method of recognizing plants using Artificial Intelligence tools like Teachable Machine and TfLite. We gathered dataset from google images and several images with varying textures and lighting conditions clicked from my smartphone camera from the Kangra forest range. There are several applications that help in recognizing the plant species using features from their leaves but we are also adding an additional feature of its care.