Video fingerprinting based on moment invariants capturing appearance and motion.
Regunathan RadhakrishnanClaus BauerPublished in: ICME (2009)
Keyphrases
- moment invariants
- object motion
- dynamic textures
- key frames
- pattern recognition
- image sequences
- zernike moments
- camera motion
- affine transformation
- feature vectors
- video sequences
- moving objects
- fourier descriptors
- optical flow
- view independent
- motion estimation
- feature points
- affine moment invariants
- appearance model
- multiscale
- object tracking
- rotation invariant
- affine invariant
- state space
- feature extraction