CVonline wiki - Motion and time sequence analysis related concepts (original) (raw)

A part of the CVonline computer vision resource summarizing the different concepts and methods commonly encountered in computer vision and image processing when processing video data.

1. Active vision
   1. Kinetic depth
   2. Image stabilization
   3. Surface reconstruction
   4. Time to contact
   5. Visual servoing
2. Appearance change analysis
3. Change and moving object detection

• 1. Background modelling
     1. Finger tracking
     2. ViBE algorithm
  2. Change detection in compressed image/video data
  3. Change detection in non-standard images: Panoramic, Omnidirectional
  4. Detection with a changing background
  5. Foreground modelling
  6. Image differencing
  7. Moving camera change detection
  8. High-speed cameras
  9. Shadow removal, Moving shadow detection
• • 1. The Stauffer and Grimson algorithm

1. Depth/Range image temporal sequence analysis
2. Image sequence fusion

• 1. Image mosaics
  2. Image stabilization
  3. Super-resolution

1. Motion field

• 1. Depth estimation
  2. Edge/Discontinuity detection
  3. Hierarchical motion field estimation
  4. Region segmentation/decomposition
     1. Motion layer/Multiple motion segmentation
  5. Particle image velocimetry

1. Motion property estimation

• 1. General motion estimation
  2. Observer motion, Egomotion estimation
  3. Periodicity estimation
  4. Planar motion estimation
  5. Instance centre of rotation
  6. Pure translation
  7. Linear motion estimation

1. Non-rigid motion analysis

• 1. Human pose and motion estimation

1. Optical flow

• 1. Affine flow
  2. The aperture problem
  3. Methods for determining optical flow
     1. Area based methods
     2. Binocular methods
     3. Contour based methods
     4. Correlation based optical flow estimation
     5. Gradient based optical flow estimation
     6. Feature based optical flow estimation
  4. Optical flow boundary, discontinuity estimation
  5. Color optical flow
  6. Optical flow field calculation
  7. Optical flow histogram
  8. Information extraction
• • 1. Egomotion estimation
       1. Epipole location
       2. Focus of expansion
       3. Obstacle detection
          1. Parking sensors
• 1. Multigrid methods
  2. Normal flow
  3. Optical flow constraint equation
  4. Optical flow smoothness constraint
  5. Range flow
  6. Scene flow/Surface motion
  7. Semantic flow / Semantic optical flow
  8. SIFT flow features
  9. Structure from optical flow

1. Sensor, Camera Motion estimation
2. Spatio-Temporal reasoning

• 1. Epipolar plane analysis
  2. Spatio-temporal corner/Interest point detector
  3. Spatio-temporal filters
     1. Singular spectrum analysis
  4. Representations
• • 1. Small motion models
• 1. Temporal shape matching
• • 1. Spatio-Temporal Relationship Match (SRM)
       1. Temporal geometric shape model matching
       2. Temporal property model matching
• 1. Temporal spatio-velocity transform

1. Structure from motion/Structure and motion

• 1. Articulated object segmentation
  2. Classical structure from motion
  3. Model-based (facial) motion capture/Model-based shape capture
  4. Multi-frame structure estimation
     1. Critical motions
     2. Initialization
     3. Euclidean reconstruction
  5. Nonlinear recursive methods
  6. Rigid bodies
• • 1. Rigid body segmentation
• 1. Rigidity constraint
• • 1. Kruppa equations
• 1. Structure consistency constraint
  2. Structure factorization
• • 1. Tomasi–Kanade factorization
       1. Factorization with uncertainty
• 1. Temporal factorization

1. Temporal event analysis

• 1. Activity analysis
     1. Temporal action segmentation
     2. Instantaneous activity recognition
     3. Long term activity
     1. Hidden Markov Model matching
     2. Hidden Semi-Markov Model matching
     3. Rule-based/Syntactic model matching
     4. Trajectory model matching
  2. Novelty detection
  3. Self-similarity matrix

1. Tracking

• 1. Articulated object tracking
  2. Binocular tracking
  3. CAMshift tracking
  4. Discontinuous events tracking
  5. Feature tracking approaches
     1. Contour tracking, Active contour tracking
     2. Appearance-based tracking
     3. Edge tracking
     4. Fiduciary marker tracking
     5. Optical flow-based target tracking
     6. Pose based tracking
     7. Feature-based tracking
     8. Template-based tracking
     9. Temporal stereo tracking
     10. Texture-based tracking
  6. Mean shift tracking
  7. Moving camera based tracking
  8. Multiple target tracking
• • 1. Feature-based tracking
       1. Optical flow analysis
       2. Template-based tracking
• 1. Tracking information fusion
• • 1. Condensation/Particle filter tracking
       1. Tracking using Bayesian belief propagation
       2. Tracking using hidden markov models
       3. Tracking using Kalman filters
• 1. Tracklets
  2. Vergence maintenance