Superpixels
This is a repo for superpixel research. It contains 4 module; they are algorithms, applications, evaluation and features.
Superpixel Algorithms
This module contains most of superpixel segmentation algorithms, such as SLIC, Turbopixel, Watershed, and so on, whose code are public avaible.
Applications
This module focuse some computer vision application based on superpixels, such as object recognition, surface rescontruction, and so on.
Evaluation
This module conatins a superpixel benchmark. There are 13 metrics for superpixel evaluation.
Features
Superpixel feature extraction is important for computer vision application, we build a standalone feature extraction module which contain common superpixel feature extract algorithms, such as SIFT, HOG.