Project Overview
Tea Leaf Disease Classifier is an AI-powered computer vision project that I developed to identify diseases in tea leaves using deep learning.
The application allows users to upload a tea leaf image and receive a model-generated prediction from eight disease categories.
My main objective was to explore how Artificial Intelligence and Computer Vision can be applied to agricultural problems, particularly plant disease identification.
I developed the project using TensorFlow, Keras, and Streamlit, combining a deep learning model with an interactive web application.
The Problem
Tea plants can be affected by different diseases, which may impact plant health, crop quality, and agricultural productivity.
Identifying these diseases early can help farmers make better decisions about crop management.
However, recognizing plant diseases through visual inspection may require experience and can become challenging when symptoms look similar.
This motivated me to explore an image-based classification approach using deep learning.
The goal was to investigate whether an AI model could recognize visual patterns associated with different tea leaf diseases.
What I Built
I developed a deep learning-based application that analyzes uploaded tea leaf images and predicts a disease category.
The project includes:
- Image Upload: Users can select a tea leaf image through the web interface.
- AI-Based Classification: A trained deep learning model analyzes the image.
- Eight Disease Categories: The system predicts among eight supported disease categories.
- Prediction Display: The application presents the model's prediction to the user.
- Interactive Web Application: Streamlit provides a simple interface for uploading images and viewing results.
How It Works
The application follows a straightforward prediction workflow.
Step 1 — Image Input
A user uploads a tea leaf image through the Streamlit interface.
Step 2 — Image Preparation
The image is prepared in the format expected by the trained model.
Step 3 — Deep Learning Prediction
The model analyzes visual characteristics in the image and produces a classification prediction.
Step 4 — Disease Classification
The system selects a predicted category from the eight supported classes.
Step 5 — Result Display
The application presents the prediction through the web interface.
This workflow demonstrates how a trained AI model can be connected to an accessible, interactive application.
Technologies Used
Python
Used for developing the machine learning workflow and application logic.
TensorFlow
Provides the deep learning framework used in the project.
Keras
Used for working with the neural network model.
Streamlit
Provides the web interface where users can upload tea leaf images and view predictions.
Computer Vision
Enables image-based analysis of tea leaves.
Deep Learning
Allows the model to learn visual patterns associated with different classification categories.
Why I Built This Project
I wanted to work on an AI project connected to a practical problem rather than focusing only on model development.
Agriculture offered an interesting application of computer vision.
Building this project allowed me to explore both sides of an AI-powered application:
- Developing an image classification model.
- Making the model accessible through an interactive web interface.
I also wanted to understand how a trained model could be used outside the development environment.
Potential Real-World Applications
The approach demonstrated in this project could contribute to future agricultural tools, such as:
- Plant disease identification systems.
- AI-assisted agricultural monitoring.
- Smart farming applications.
- Crop-health analysis tools.
- Educational demonstrations of agricultural computer vision.
These are potential applications. Real agricultural use would require further evaluation with representative field images and expert validation.
What I Learned
This project helped me strengthen my understanding of:
- Applying deep learning to image classification.
- Using TensorFlow and Keras for AI model development.
- Working with image-based prediction workflows.
- Connecting a trained model to a web interface.
- Building interactive machine learning demonstrations with Streamlit.
- Exploring the applications of AI in agriculture.
One of the most valuable parts of this project was learning how to turn a machine learning model into an application that others can interact with.
Project Links
- Source Code: View on GitHub
- Live Demo: Try the Tea Leaf Disease Classifier
- LinkedIn Post: View my original project announcement
This project is part of my learning journey in Artificial Intelligence, Deep Learning, Computer Vision, and their practical applications.
