Skip to content

Latest commit

 

History

History
executable file
·
78 lines (54 loc) · 2.01 KB

README.md

File metadata and controls

executable file
·
78 lines (54 loc) · 2.01 KB
Raw

plant_diseases_recognition

This project aims to develop a model (CNN) for recognizing plant diseases using deep learning. The model is trained on the PlantVillage dataset, and it classifies different plant diseases based on images of leaves.

Requirements

  • Python 3.12.3
  • TensorFlow 2.17.0
  • Poetry for dependency management.

Installation

  1. Clone the repository:

    git clone https://github.com/MilleXi/plant_diseases_recognition.git

  2. Clone the PlantVillage dataset:

    git clone https://github.com/gabrieldgf4/PlantVillage-Dataset.git
    • Note: After downloading the dataset, delete the x_Removed_from_Healthy_leaves folder and the .git folder inside the PlantVillage-Dataset directory.

  3. Install the required Python dependencies:

    poetry install

Configuration

You can modify the configuration settings by editing the config.py file located in the config folder. This file contains various parameters related to model training and dataset paths.

Training

To train the model, run the following command in the terminal:

python train.py

Evaluating

To evaluate the model, run the following command in the terminal:

python evaluate.py

The model will be trained and evaluated on the PlantVillage dataset, and the training output, including logs and model checkpoints, can be found in the output folder.

Feature maps

If you want to see the feature maps, run the following command in the terminal:

python get_feature_maps.py

You can find the pictures in the output folder.

Visual Interface

To open the visual interface, run the following command in the terminal:

python gradio_interface.py

Output

After you run the above code, you can find the following in the output folder:

  • Checkpoints
  • Feature maps
  • Logs
  • Best model in the 'models' folder
  • Other Visualizations

License

This project is licensed under the Apache-2.0 license - see the LICENSE file for details.