Background subtraction by probabilistic modeling of patch features learned by deep autoencoders.
Jorge García-GonzálezJuan Miguel Ortiz-de-Lazcano-LobatoRafael M. Luque-BaenaEzequiel López-RubioPublished in: Integr. Comput. Aided Eng. (2020)
Keyphrases
- background subtraction
- probabilistic modeling
- moving objects
- motion detection
- probabilistic model
- video surveillance
- object detection
- feature extraction
- feature vectors
- gaussian mixture model
- image sequences
- foreground detection
- foreground objects
- post processing
- bayesian inference
- background model
- background modeling
- background image
- denoising
- image patches
- test images
- appearance model
- feature space
- human detection
- preprocessing
- bayesian networks
- training data
- similarity measure