Classification and Pesticide Detection in Fruits

Abstract

Pesticides being present in fruits is an increasing concern for people all over the world. By working with the customer, this work aims to resolve the issue. The consumer should learn if the chosen fruits are suitable for consumption. The primary goal of this research is to use a sensor to calculate the normalized differential vegetation index in order to detect the levels of pesticides in fruit. The work is divided into two parts here. Utilizing Convolutional Neural Network (CNN), the first section identifies various fruits. Here, CNN trains a set of visual attributes for classifying fruits and their freshness, including color, shape, and texture. Module 2 detects pesticides in fruits by calculating NDVI with an IR sensor, gas sensor, soil moisture sensor, temperature and humidity sensor, and comparing the results. The Arduino's program specifies the output on the display screen. The screen will show the detection information. And a plot of the resulting graph is made.