Multi-feature fusion based background subtraction for video sequences with strong background changes.
Zhenkun HuangRuimin HuThierry BouwmansShihong ChenPublished in: ICIP (2017)
Keyphrases
- foreground objects
- background subtraction
- feature fusion
- video sequences
- background model
- moving objects
- dynamic background
- foreground detection
- image sequences
- background modeling
- dynamic scenes
- background image
- moving camera
- object detection
- gaussian mixture model
- motion detection
- video surveillance
- feature extraction
- moving object detection
- pixel level
- multiple features
- video frames
- foreground regions
- object segmentation
- object tracking
- post processing
- motion estimation
- machine learning
- computer vision
- multiresolution
- human motion
- gait recognition
- fusion method
- mixture model
- appearance model
- three dimensional