Smart Harvest: Web-Integrated Ripeness Detection for Apples with CNN Algorithm

Abstract

This study presents "Smart Harvest," a web-based system that aids farmers, sellers, or buyers in identifying the maturity level of apples based on skin colour using a Convolutional Neural Network (CNN) algorithm. Traditional methods, reliant on human labour, suffer from subjectivity and inconsistency in evaluating fruit maturity. Our system offers an automated, objective, and reliable alternative to address this. By analyzing skin colour, Smart Harvest classifies apples into two primary maturity levels: raw and ripe. Through rigorous training and testing phases, the system has demonstrated remarkable efficiency, achieving an impressive average prediction accuracy of 93-94%. This paper details the development and deployment of Smart Harvest, showcasing its potential to enhance agricultural productivity by providing a user-friendly, web-based tool for accurate ripeness detection. The implementation of such a system not only promises to standardize the maturity assessment process but also aims to optimize harvesting schedules, reduce labour costs, and minimize waste, thereby contributing significantly to the agricultural sector's sustainability and profitability.