Background modeling using adaptive pixelwise kernel variances in a hybrid feature space.
Manjunath NarayanaAllen R. HansonErik G. Learned-MillerPublished in: CVPR (2012)
Keyphrases
- feature space
- background modeling
- pixel wise
- background subtraction
- background model
- foreground objects
- gaussian mixture model
- object detection
- subspace learning
- moving objects
- mean shift
- video surveillance
- spatial information
- feature selection
- image representation
- object tracking
- feature vectors
- dimensionality reduction
- appearance model
- template matching
- frame rate
- pixel level
- training set
- principal component analysis
- feature extraction
- high dimensional
- low dimensional
- support vector
- object recognition
- sparse coding
- data points