Feature detection and matching are an integral part of computer vision, and are fundamental for many applications. Finding feature correspondence between images is a fundamental part of many applications. Feature detection and matching is comprised of three stages.

  1. Feature detection using one of the following methods: Features from Accelerated Segment Test (FAST) Detector, Harris Corner Detector, or Shi-Tomasi Corner Detector.
  2. Feature description using one of the following methods: Binary Robust Invariant Scalable Keypoints Descriptor (BRISK) or Fast Retina Keypoint Descriptor (FREAK).
  3. Feature matching.