VLFeat

About VLFeat

VLFeat is an open source library focused on computer vision, providing efficient implementations of popular algorithms for image understanding, feature extraction, and matching. It is suitable for researchers, students, and developers working in computer vision, with native C code and convenient MATLAB interfaces, as well as command-line and Octave support. VLFeat's robust library includes algorithms like SIFT, MSER, Fisher Vectors, clustering, superpixels, SVMs, and more, and it is cross-platform (Windows, Mac, Linux) with extensive documentation and tutorials.