Video Segmentation via a Gaussian Switch Background Model and Higher Order Markov Random Fields.
Martin RadolkoEnrico GutzeitPublished in: VISAPP (1) (2015)
Keyphrases
- video segmentation
- markov random field
- higher order
- background model
- background subtraction
- foreground objects
- video sequences
- gaussian mixture model
- video frames
- moving objects
- motion detection
- graph cuts
- segmentation method
- energy function
- segmentation algorithm
- random fields
- maximum a posteriori
- natural images
- parameter estimation
- conditional random fields
- pairwise
- video analysis
- maximum likelihood
- pixel level
- moving target
- image segmentation
- higher order cliques
- video data
- video surveillance
- detection algorithm
- object detection
- object segmentation
- feature vectors
- image sequences
- moving camera
- saliency map
- dynamic scenes
- appearance model
- object tracking
- image features
- state space