Arpit Aggarwal
SLIC stands for Simple Linear Iterative Clustering. This is a method used for Superpixel generation. Superpixels help in clustering pixels of similar color and texture together and instead of working with thousands of pixels, work with superpixels can be done as pixels within the superpixel would be similar to one another on the basis of color and texture. SLIC algorithm differs from K-Means in three ways: we work with images and keep the spatial information intact, we confine pixels to be within a superpixel only if they are in a 2S x 2S area where S is a parameter and they use LAB space instead of RGB space because color distance in RGB space is not meaningful.
The dataset used is provided in notebooks folder. Five images are used on which the SLIC algorithm for Superpixel generation is tested.
The results obtained after applying SLIC algorithm for Superpixel generation are shown below:
To run the jupyter notebooks, use Python 3. Standard libraries like Numpy, Scipy and OpenCV are used.
The following links were helpful for this project: