Smart Prediction of Shelf Life and Tomato Sorting Using Deep Learning

Abstract

Agriculture has been essential throughout history, sustaining communities and providing food worldwide. Over time, farming methods have evolved with new ideas and scientific advancements. Today, technology plays a crucial role in improving farming practices. Deep learning and computer vision are two such technologies that help farmers by providing smart solutions like crop monitoring and smart irrigation system. Tomatoes are a significant crop with versatile uses and high consumer demand. However, traditional methods of sorting tomatoes are labor-intensive, time consuming and prone to errors, leading to delays in the supply chain. To address this issue, the proposed system introduces a object detection model to enhance tomato sorting processes. In addition to software solutions, the proposed system aims to develop an automated sorting prototype equipped with specialized components and technology. This proposed system is capable of dynamically sorting tomatoes based on ripeness and quality. The proposed system carried out by using two models related to YOLOv8 types for determining the shelf life of tomato in smart way. In the present Food security era, this research delivers to sensitize the commercial tomato growers through sending some sensor base signals or warnings to concern people through IoT communication. As a tomato undergoes assessment by the computer, the prototype responds accordingly, minimizing manual intervention and reducing errors. By integrating sophisticated technology with practical machinery, our objective is to optimize agriculture methods. This proposed system aims to showcase the transformative potential of technology in agriculture by enabling efficient sorting of tomatoes, leading to increased productivity, resource efficiency, and food security.