A 3.0μW@5fps QQVGA Self-Controlled Wake-Up Imager with On-Chip Motion Detection, Auto-Exposure and Object Recognition.
Arnaud VerdantWilliam GuicqueroNicolas RoyerGuillaume MoritzSébastien MartinFlorent LepinSylvain ChoisnetFabrice GuellecBenoît DeschampsSylvain ClercJérôme ChossatPublished in: VLSI Circuits (2020)
Keyphrases
- motion detection
- object recognition
- high speed
- low cost
- detection algorithm
- video surveillance
- image sequences
- background subtraction
- motion estimation
- image sensor
- focal plane
- computer vision
- real time
- surveillance system
- moving target
- frame rate
- object tracking
- object detection
- power consumption
- infrared
- image features
- unsupervised learning
- d objects
- video camera
- flow field
- image processing
- signal processing
- low power
- multi class
- thresholding method
- maximum likelihood
- image segmentation
- motion analysis
- three dimensional
- appearance model
- reinforcement learning