Real-Time Flora Species Identification Using Unmanned Aerial Vehicle

Remote sensing is used in drones, and Unmanned Aerial Vehicles (UAVs) current trends, and this technology can provide unparalleled benefits that change lives in humanity. Drones are used in many fields such as the military, research, agriculture, and recreation. This technology can not only reduce test time, but also provide many benefits such as providing a real-time live video test site that can help the user analyze site planning and speed up all activities. However, green monitoring using a sensor remotely has its challenges with regard to image processing and data. Somehow, previous research has suggested a number of possible algorithms that can be used to process the data collected after processing. However, most of the algorithms require a certain sensor to get the result. In this work, mini sensors and processors like raspberry pi for the flight controller are used. The purpose of this paper is to identify and validate an appropriate algorithm for processing plant images. In turn, using these algorithms is used to identify trees and natural vegetation. This study uses the data collected from a variety of sources using a consumer camera and process using convolutional neural networks (CNN) models to give appropriate results. Proposed model is trained to identify the mango leaves and the model gave 95% accurate results.