A new strategy based on adaptive mixture of Gaussians for real-time moving objects segmentation.
Carlos CuevasNarciso GarcíaLuis SalgadoPublished in: Real-Time Image Processing (2008)
Keyphrases
- mixture of gaussians
- background subtraction
- moving objects
- real time
- background modeling
- background model
- motion segmentation
- video surveillance
- maximum likelihood estimation
- foreground objects
- image sequences
- video sequences
- image segmentation
- moving camera
- level set
- segmentation algorithm
- segmentation method
- moving target
- motion detection
- object tracking
- gaussian mixture model
- density function
- pixel level
- dynamic scenes
- density estimation
- detection algorithm
- gaussian mixture
- medical images
- object detection
- gaussian model
- class probabilities
- gaussian distribution
- mixture model
- em algorithm
- expectation maximization
- multi class
- optical flow
- training set
- decision trees