SLIC-Simple-Linear-Iterative-Clustering
Author: Shantam Bajpai
Course: CMSC 828I: Advanced Visual Learning and Recognition
Description of the Project
In this project the Simple Linear Iterative Clustering algorithm has been implemented from sratch. In addition Kmeans , KMeans on 5D RGBXY space and KMeans with weighted distances has been implemented from scratch to compare the pixelwise and superpixel outputs of SLIC and Classical Unsupervised Clustering approaches.
Outputs
The outputs can be views in the IPython notebook that has been provided. Tested the 3 algorithms for various values of k (Number of clusters).
Libraries Used
- OpenCV ( Primary for Loading Image data)
- Numpy
Research Paper Reference
Simple Linear Iterative Clustering: https://www.iro.umontreal.ca/~mignotte/IFT6150/Articles/SLIC_Superpixels.pdf