Welcome to a fast and interactive web app for image processing using filters like Sobel, Gaussian, and Median! Built with Python, NumPy, Pillow, and accelerated with Cython. Deployed live using Streamlit Cloud.
This project aims to showcase the potential of High Performance Computing (HPC) in image processing through a comparative study of three implementations: pure Python, NumPy, and Cython. Each version applies the same filters and reports the processing time, enabling a visual and performance-based evaluation of optimization techniques.
Through this demonstration, we highlight how computing performance improves by moving from interpreted code to compiled extensions — a key idea behind HPC.
- 🌀 Apply Sobel, Gaussian, and Median filters to images
- 📈 Visualize and compare performance across Python, NumPy, and Cython
- 📷 Upload your own images to test
- ☁️ Fully deployable on Streamlit Cloud
- ⚡ Performance boosted using Cython extensions
Python🐍NumPy📦Pillow🖼️Cython⚡Streamlit🌐Matplotlib📊SciPy🔬Setuptools🛠️Pathlib🧱
To run the app locally:
git clone https://github.com/Oscaretz/Processing_Image
cd processing_image
pip install -r requirements.txt
streamlit run app.pyThis project supports Cython to accelerate heavy operations. To compile Cython modules:
- Add your
.pyxfile (e.g.,process_image_cython.pyx) - Create or update
setup.py:
from setuptools import setup
from Cython.Build import cythonize
setup(
ext_modules=cythonize("process_image_cython.pyx")
)- Compile it:
python setup.py build_ext --inplaceTo compare performance:
python unit_test_python.py
python unit_test_numpy.py
python unit_test_cython.pyThis app is already deployed on Streamlit Cloud! 🔗
You can also deploy your own fork by pushing it to your GitHub and connecting it to Streamlit Cloud.