Unsupervised Learning for Target Tracking and Background Subtraction in Satellite Imagery.
Jonathan S. KentCharles C. WamsleyDavin FlateauAmber FergusonPublished in: CoRR (2021)
Keyphrases
- background subtraction
- target tracking
- satellite imagery
- unsupervised learning
- remote sensing
- motion detection
- moving target
- satellite images
- image sequences
- moving objects
- data fusion
- kalman filter
- surveillance system
- video surveillance
- object detection
- supervised learning
- gaussian mixture model
- mixture model
- dynamic scenes
- mean shift
- particle filter
- video camera
- appearance model
- moving camera
- expectation maximization
- object tracking
- dimensionality reduction
- multi sensor
- semi supervised
- object recognition
- feature selection
- machine learning
- data association
- post processing
- high resolution
- active learning
- model selection
- motion model
- video sequences
- multispectral
- high quality