Apppublic

LEGENDFTW/image-segmentation-explorer

sourceHugging Faceupdated 4mo agoView on Hugging Face

0likes

App README

Image Segmentation Explorer

An interactive Streamlit app for exploring three classic image segmentation algorithms. Built as a course assignment for an image analysis module. Link- https://huggingface.co/spaces/LEGENDFTW/image-segmentation-explorer

What it does

—Upload your own image or choose from 3 built-in samples
—Choose between Thresholding, K-means, and Watershed
—Adjust parameters with sliders and see results instantly
—Side-by-side original / segmented / overlay comparison
—Diagnostics: segment count, cluster sizes, pixel histograms
—Theory tab explaining how each method works

Run locally

bash

pip install -r requirements.txt
streamlit run app.py

Known limitations

—Watershed can be slow on large images (resize to <600px recommended)
—K-means result varies slightly between runs (random initialisation)
—Thresholding fails on images with uneven lighting unless Adaptive mode is used