- Version:
- 0.9.15
- Copyright © 2007-12 Andrea Vedaldi and Brian Fulkerson
- Copyright © 2012 The VLFeat Team
The VLFeat C library implements popular computer vision algorithms, with a special focus on feature detectors and descriptors. Applications range from wide-baseline matching to image classification and object detection. VLFeat aims to be simple to use, portable, and thoroughly documented.
Contents
- Introduction
- Computer vision algorithms
- Covariant image frames detector
- Gaussian Scale Space
- Dense Scale Invariant Feature Transform (DSIFT)
- Maximally Stable Extremal Regions
- K-means clustering
- Integer K-means (IKM)
- Hierarchical Integer K-means (HIKM)
- Agglomerative Information Bottleneck (AIB)
- KD-trees and forests
- Homogeneous kernel map
- PEGASOS SVM solver
- Simple Linear Iterative Clustering (SLIC)
- Quick shift image segmentation
- Support functionalities
- Design
- Developer manual
- Glossary
Introduction
The VLFeat project is a collaborative effort started in 2007 by Andrea Vedaldi and Brian Fulkerson at the University of California at Los Angeles and now developed by the VLFeat Team. The project includes the development of a C library, implementing a number of popular computer vision algorithms, as well as wrappers for scripting languages such as MATLAB.
The reference documentation of the algorithms implemented in VLFeat is contained in this manual, describing the C API. Other documentation (for example the help built into VLFeat MATLAB functions) does not contain a similarly complete and detailed description.
Using the library from a C program is for the most part obvious from the examples included in the descriptions of the various algorithms (see the Contents for a list). The general design of the library (including for example the handling of errors and the ownership conventions for allocated objects) is descried in Design. Further information for the library developers is described in the Developer manual.