Precision-Spraying-for-Agriculture-mobile-Robot

This project is the HEC funded project, the aim of Precision Agriculture Spraying robot is to develop a robotic solution for precisely spraying application for different crops using Machine learning and Deep learning algorithms.

DATASET FOR WEED AND PLANT

we used TobSet for our model training, testing and evaluation. dataset link: https://github.com/mshahabalam.

CNN model training

Precision_Spraying_training_CNN.ipynb is used for training our model on TobSet dataset.

CNN model testing

Precision_Sprayer_testing_CNN.ipynb is used to test our model on different unseen images some of the results are shown

Real time spraying and detection of tobacco plants

Real_time_Spraying .ipynb is used to implement our model in real time using webcam, in this file python language is interfaced with Arduino to activate the nozzle solenoid valve. In this code we have splitted our webcam video into two halfs, right and left video. if the tobacco plant is detected in right video so right nozzle should activate and if the tobacco plant is detected in left video so left solenoid should activate.

Image processing to detect tobacco plant

Segmentation and Detection.ipynb file contains the code for detecting only tobacco plant using only image processing techniques. results are: