State space approach to constrained recursive deconvolution of a noisy image sequence.
Michael S. MortMandyam D. SrinathPublished in: ICASSP (1988)
Keyphrases
- state space
- image sequences
- spatio temporal
- least squares
- dynamic programming
- heuristic search
- state variables
- reinforcement learning
- video sequences
- motion field
- particle filter
- optical flow
- dynamical systems
- denoising
- motion model
- image restoration
- markov decision processes
- three dimensional
- continuous time markov process
- optimal policy
- consecutive images
- motion estimation
- image frames
- noisy data
- planning problems
- search space
- computer vision
- action space
- reinforcement learning algorithms
- motion analysis
- structure from motion
- motion segmentation
- missing data
- optic flow
- background subtraction
- outdoor scenes
- consecutive frames
- high resolution
- blind deconvolution
- feature points
- fir filters
- goal state
- stochastic domains
- camera motion