DIY Farm Drone(NDVI)

April 2022

Purpose

The Drone will trace through whole farm and capture images from Raspberry Pi and NoIR (InfraRed) Camera. Then using NDVI it will calculate the healthy and unhealthy plants from the top view of the crop. Further using Deep Learning it will predict if fruit has any diseases or not and also check the ripeness of the fruit.

How it Works:

NDVI (Normalized Difference Vegetation Index) works on the principle that leafs with photosynthesis and without photosynthesis reflect Infra Red Rays differently. We take help of theis principle to differentiate between helathy and weak plants.

Further using Deep Learning we track the fruits from plants and try to find out if it has any diseases or not.The model is trained on set of diseased and healthy fruit images of Mango and Apple

NDVI Principle

VIS- Visible Light NIR- Near InfraRed Light

Drone Specification

The drone type is a FPV Racing drone and a 3D printed holder will support the Multispectral camera(NOIR Camera) and Raspberry Pi from below to take images of the farm.

The drone consists of the following components :

• 2400KV BLDC motors from EMAX
• SP Racing F3 Flight Controller
• 30Amp ESCs with BlHeli support
• 2400MaH LiPO battery
• 2.4Ghz RC Controller 6 Channels

And the software used is Cleanflight and BetaFlight

Images