Image Denoiser

A small Signals & Systems project that demonstrates low-pass filtering for noise removal in images.

Features

• CLI mode for quick experiments and batch noise sweeps.
• Flask web frontend for uploading images, tuning filters, and previewing results.
• Saved outputs for the original image, noisy image, and each filtered result.

Quick start

Create and activate the virtual environment (optional if you already have one):

python -m venv .venv
Set-ExecutionPolicy -Scope CurrentUser -ExecutionPolicy RemoteSigned (bypass if needed)
& .venv\Scripts\Activate.ps1
pip install -r requirements.txt

Run the demo on a single image (saves results to results/):

python image_denoiser.py --input tsghost2.jpg --outdir results --noise-std 0.1 --filters gaussian fft --no-show

Batch noise example:

python image_denoiser.py --input tsghost2.jpg --batch-noise 0.01 0.05 0.1 0.2 --outdir experiments --no-show

If --input is omitted, the script uses the scikit-image camera test image.

Web frontend

Start the browser UI from the project venv:

& .venv\Scripts\python.exe web_app.py

Then open http://127.0.0.1:8501/ in your browser.

In the web UI you can:

• Upload an image or pick one already in the project folder.
• Set noise strength, kernel size, sigma, and FFT cutoff.
• Choose which filters to run.
• View SNR scores and download the generated PNG outputs from web_results/.

Troubleshooting

• If python web_app.py fails with No module named 'flask', run the app with the venv interpreter: & .venv\Scripts\python.exe web_app.py.
• If dependencies are missing, install them with & .venv\Scripts\python.exe -m pip install -r requirements.txt.