Tomato Quality Classification using hybrid Swin Transformer

Project Code :TCMAPY1804

Objective

The objective of this project is to develop an automated system for classifying tomato quality into distinct categories: Unripe, Ripe, Old, and Damaged, using deep learning techniques. By leveraging a combination of Convolutional Neural Networks (CNN), ResNet, and a hybrid Swin Transformer with Random Forest (RF), the project aims to improve the accuracy of tomato quality detection. This system intends to streamline quality control processes in agriculture, ensuring efficient distribution of fresh produce while reducing human error. Ultimately, the project aims to enhance the decision-making process in the agricultural supply chain through reliable, real-time tomato quality assessment.

Abstract

Tomato quality classification is a critical task in the agricultural industry, aimed at ensuring the efficient distribution of fresh and quality produce to consumers. In this study, we propose a deep learning-based approach for classifying tomatoes into multiple quality categories, including Unripe, Ripe, Old, and Damaged. The dataset used for this classification is sourced from Kaggle's "Tomatoes Dataset," which contains labeled images of tomatoes in various stages of ripeness and health.

Keywords: Tomato quality classification, multiclass classification, CNN, ResNet, Swin Transformer, Random Forest, deep learning, image classification, agricultural supply chain, quality assessment.

NOTE: Without the concern of our team, please don't submit to the college. This Abstract varies based on student requirements.