This project implements an image dehazing technique using the AOD-Net architecture.
readme.md
: Instructions for setting up and running the project.requirements.txt
: Contains all necessary Python libraries.dehazing-code.ipynb
: Jupyter notebook with the full dehazing code.dehazing.py
: Python code to run the dehazing UI on Windows OS.run_cvip_ui_picker.bat
: Batch file to launch the dehazing UI on Windows.run.py
: Python script to run the UI on any operating system.test_dataset/
: Folder containing hazy image test data.AOD_Net_reg.h5
: Saved model in Hierarchical Data Format (HDF).
Follow these instructions to set up your environment to run the project.
Ensure you have Python 3.6 or higher installed on your machine. Install all the required dependencies with one of the following methods:
pip install -r requirements.txt
Alternatively, you can install the dependencies directly using:
pip install opencv-python numpy tensorflow matplotlib scikit-image keras tkinter PIL
- Double-click the run_cvip_ui_picker.bat file to start the application.
- Wait for the app to open, and then upload an image from the directory using the UI.
On macOS or Linux:
-
Open a terminal.
-
Execute the following command:
python run.py
- To dehaze a different image, change the image path in the img variable inside run.py. For example: img = cv2.imread('file_path')
Note: Keep the command prompt open while the app is running.