Tutorial Geometric Vision

Geometric Computer Vision Tutorial

A major component of 3D computer vision is the study of epipolar geometry, the geometry of two views. BoofCV provides many standard algorithms for mathematically describing features across two views and more. Below is a list of these algorithms. The API is still being worked on and a more detailed tutorial will be written later on.

Recommend Reading:

• Yi Ma, et. al., "An Invitation to 3-D Vision"
  • Algebraic approach to 3D vision.
  • Easy to use as a reference.
• R. Hartley, and A. Zisserman, "Multiple View Geometry in Computer Vision"
  • Geometric approach and the most popular book on this subject.
  • More content and good algorithms.
  • Mediocre index and tendency to not define terms in some chapters.

Algorithm List

• Fundamental Matrix 8 Points (Linear)
• Fundamental Matrix 7 Points (Linear)
• Fundamental Optimization
  • Sampson Error
  • Epipolar Error
• Homography 4 Points (Linear)
• Homography Optimization
  • Sampson Error
  • Transfer Error
• Efficient PnP 4-Point
• Linear 6 Point Pose
• Linear Pixel Depth
• Triangulation Geometric
• Triangulation Linear
• Triangulation Optimization
  • Sampson Error
  • Euclidean Error
• Decompose Essential
• Decompose Homography
• Bundle Adjustment
  • Dense (inefficient/slow)
• Stereo Rectification
  • Calibrated
  • Uncalibrated

Examples

• Compute Fundamental Matrix

Related

• Zhang 99 Camera Calibration
• Various different feature trackers and detectors