TeaLeafNet: A Hybrid CNN-Boosting Approach for Tea Leaf disease Classification

Tea Leaf Disease Classification

Overview

This repository contains code and resources for a machine learning project aimed at detecting tea leaf diseases using convolutional neural networks (CNNs). The project involves the use of a dataset comprising images of tea leaves under various conditions, including both healthy and diseased leaves.

Objectives

Dataset Preparation:

• Utilized a dataset of tea leaf images sourced from multiple providers.
• Sorted the dataset based on different factors, including the type of disease.

CNN Investigation:

• Explored multiple convolutional neural network architectures for visual recognition of tea leaf diseases.

Model Development:

• Proposed a comprehensive machine learning model for assessment.
• The model includes an encoder-decoder CNN-Boosting and incorporates two pre-trained CNN models with transfer learning (TL).

Performance Evaluation:

• Evaluated the performance of the CNN-Boosting model using fundamental evaluation measures, such as accuracy, precision, recall, and F1 score.

• Dataset: https://data.mendeley.com/datasets/j32xdt2ff5/1