🍃 Leaffliction - Leaf Disease Recognition using Computer Vision

Overview

Leaffliction is a deep learning project aimed at identifying diseases in leaves through computer vision techniques. The dataset is from PlantVillage. It leverages image processing, augmentation, and Convolutional Neural Networks using Python libraries such as FastAI, OpenCV, and PyTorch.

Features

• Data Preprocessing: Includes image augmentation and transformation for dataset enhancement.
• Model Training: Utilizes FastAI with a VGG19 model for training on augmented datasets.
• Disease Classification: Employs classification techniques to identify various leaf diseases.
• Image Analysis Tools: Features tools for image analysis and visualization.

Installation

pip install -r requirements.txt

Usage

Training the Model

To train the model, use the following command with the path to your training dataset:

python train.py /path/to/dataset

Predicting Diseases

For disease prediction on new leaf images, use:

python predict.py /path/to/image_or_directory

Image Augmentation

To augment images in your dataset, execute:

python Augmentation.py /path/to/images

Image Transformation

For applying various transformations to your images:

python Transformation.py /path/to/image_or_directory

Data Visualization

To visualize the distribution of classes in your dataset:

python data_visualization.py /path/to/dataset

Acknowledgments

Made in collaboration with Walter / Jeremy