This project implements a graph-based image segmentation algorithm, dividing an image into regions using edge-weighted graph analysis.
Image segmentation is the process of partitioning an image into regions with internal coherence and boundary significance.
- Graph construction from image pixels using 8-connected neighbors
- Edge weighting based on intensity differences
- Gaussian filter preprocessing to reduce noise
- Region merging using an adaptive threshold predicate
- Support for both grayscale and RGB images
- Segmented image visualization with color-coded regions