max planck institut
mpii logoMinerva of the Max Planck Society

Computer Vision and its Applications in Graphics

Advanced lecture: Computer Vision and its Applications in Graphics – Summer Semester 2010

Thorsten Thormählen

This lecture covers advanced topics in the field of computer vision and computer graphics. It will be given in English. The target audience are students in computer science or related fields. Basic knowledge about 3D geometry, image processing, and computer graphics is beneficial. In order to be admitted to the final oral exam at the end of the semester, five homework assignments must be passed.

Time: Mondays, 16:15-17:45
First lecture: Monday, April 19, 2010
Room: Building E1.4 (Max-Planck-Institut), Room 024
HISPOS: Lect. No. 44886

The lecture is structured as follows:

  1. Scene representation
    • Scene elements: lighting, geometry, cameras
    • Basic animations
    • Homework: Generate your own animated scene using Collada

  2. The history of visual special effects
    • Visual illusions
    • Matte painting

  3. Advanced camera models
    • Finite cameras
    • Projective cameras
    • Affine cameras
    • Other cameras
    • Lens distortions
    • Homework: Camera matrix decomposition

  4. Foundations of camera motion estimation
    • Feature detection
    • Feature tracking
    • 2D homography
    • Fundamental matrix
    • Trifocal tensor
    • Applications

  5. Camera motion estimation
    • N-view approaches
    • (Sparse) bundle adjustment
    • Self calibration
    • Applications
    • Homework: Camera motion estimation with Voodoo Camera Tracker and Blender

  6. 3D reconstruction
    • Stereo matching
    • Multi-view stereo
    • Applications

  7. 3D scanning
    • The principle
    • Calibration
    • Registration
    • Applications

  8. Animations
    • Keyframe animations
    • Morphing
    • Skeletons and skinning
    • Statistical models
    • Homework: Animate a 3D scan of yourself

  9. Image-based motion capture
    • Segmentation
    • Edge detection
    • Human motion estimation
    • Applications
    • Homework: Estimate the "Eigenfaces" from a number of photographs of human faces

Additional material for participants can be found on this website. It is accessible only from within the university or by correct authentication.