Abstract: Fast moving objects may be monitored by a video unit under a variety of lighting conditions. In a bootstrap phase, a model database is established for each of a plurality of sampling points, each model being a combination of color of the video image, respective optimum exposure, and respective lighting condition. Then in an image capture phase the video unit compares captured color and optimum exposure of the sampling point, with the models of color and respective optimum exposure in the model database, to find a best match lighting condition model in the model database. The video unit then adjusts an exposure setting to the most frequently occurring value for exposure among the other sampling points, if lighting conditions for the other sampling points are within a threshold range of the lighting condition of the best match model in the model database.

Abstract: Transient connected regions are tracked in a video sequence, marking them as foreground layer or background layer when the transient regions become stable. A stack of background/foreground Gaussian mixer models is maintained for each pixel. The decision to mark a stable region as background layer or a new or existing foreground layer is done by matching the region with each model in the model stack. If the new region matches an existing model, then the layers above the matched layers are purged, or else the new region is pushed as a new foreground layer with a new model.

Abstract: A distributed system is disclosed for mining, correlating and analyzing locally obtained traffic video data. The system comprises a central controller configured to correlate traffic counting meta data gathered at a thoroughfare, to transmit control information to specify one or more traffic counting events to be monitored by at least one video unit located at the thoroughfare and to receive traffic counting meta data from the at least one video unit. The at least one video unit located at the thoroughfare, is configured to process a video stream from a video camera while monitoring the traffic counting events at the thoroughfare specified by the central controller, to analyze and evaluate the traffic counting event, to encode traffic meta data characterizing the analysis, and to transmit the meta data to the central controller. Further meta data can be acquired by mining the meta data history or any captured raw video footage.

Abstract: A distributed system for traffic monitoring and event detection includes a plurality of video units within viewing distance of a thoroughfare to be monitored. Each video unit processes a video stream while monitoring events at the thoroughfare, identifies and analyzes events, encodes traffic meta data characterizing the analysis, and transmits the meta data to a central controller. The central controller stores a geographic map representing the thoroughfare, correlates the meta data received from the video units, and forms a composite representation of the geographic map and the correlated meta data. The central controller replays previously stored composite representations of the geographic map and the correlated meta data, which were monitored under mutually different event conditions. The central controller is also configured to allow changes to previously stored composite representations of the geographic map and to the correlated meta data for the purpose of understanding the impact of change.

Abstract: A parking management system includes a parking facility management controller configured to allocate parking spaces at a parking facility as at least one of customer reserved spaces, handicap spaces, electric vehicle station spaces, premium spaces or unreserved spaces, the controller further including a communications unit configured to transmit the allocation of parking spaces to at least one video unit located at the parking facility. The video unit located at the parking facility, includes a video camera and video frame processing logic to process a video stream from the video camera while monitoring a parking space allocated by the controller, identify an event occurring at the allocated parking space, and encode a low bandwidth message characterizing the event. The at least one video unit includes a communications unit configured to transmit the low bandwidth message over low bandwidth communication links to the management controller.

Abstract: Fast moving objects may be monitored by a video unit under a variety of lighting conditions. In a bootstrap phase, a model database is established for each of a plurality of sampling points, each model being a combination of color of the video image, respective optimum exposure, and respective lighting condition. Then in an image capture phase the video unit compares captured color and optimum exposure of the sampling point, with the models of color and respective optimum exposure in the model database, to find a best match lighting condition model in the model database. The video unit then adjusts an exposure setting to the most frequently occurring value for exposure among the other sampling points, if lighting conditions for the other sampling points are within a threshold range of the lighting condition of the best match model in the model database.

Abstract: Transient connected regions are tracked in a video sequence, marking them as foreground layer or background layer when the transient regions become stable. A stack of background/foreground Gaussian mixer models is maintained for each pixel. The decision to mark a stable region as background layer or a new or existing foreground layer is done by matching the region with each model in the model stack. If the new region matches an existing model, then the layers above the matched layers are purged, or else the new region is pushed as a new foreground layer with a new model.

Abstract: A distributed system is disclosed for mining, correlating and analyzing locally obtained traffic video data. The system comprises a central controller configured to correlate traffic counting meta data gathered at a thoroughfare, to transmit control information to specify one or more traffic counting events to be monitored by at least one video unit located at the thoroughfare and to receive traffic counting meta data from the at least one video unit. The at least one video unit located at the thoroughfare, is configured to process a video stream from a video camera while monitoring the traffic counting events at the thoroughfare specified by the central controller, to analyze and evaluate the traffic counting event, to encode traffic meta data characterizing the analysis, and to transmit the meta data to the central controller. Further meta data can be acquired by mining the meta data history or any captured raw video footage.

Abstract: A distributed system for traffic monitoring and event detection includes a plurality of video units within viewing distance of a thoroughfare to be monitored. Each video unit processes a video stream while monitoring events at the thoroughfare, identifies and analyzes events, encodes traffic meta data characterizing the analysis, and transmits the meta data to a central controller. The central controller stores a geographic map representing the thoroughfare, correlates the meta data received from the video units, and forms a composite representation of the geographic map and the correlated meta data. The central controller replays previously stored composite representations of the geographic map and the correlated meta data, which were monitored under mutually different event conditions. The central controller is also configured to allow changes to previously stored composite representations of the geographic map and to the correlated meta data for the purpose of understanding the impact of change.

Abstract: A parking management system includes a parking facility management controller configured to allocate parking spaces at a parking facility as at least one of customer reserved spaces, handicap spaces, electric vehicle station spaces, premium spaces or unreserved spaces, the controller further including a communications unit configured to transmit the allocation of parking spaces to at least one video unit located at the parking facility. The video unit located at the parking facility, includes a video camera and video frame processing logic to process a video stream from the video camera while monitoring a parking space allocated by the controller, identify an event occurring at the allocated parking space, and encode a low bandwidth message characterizing the event. The at least one video unit includes a communications unit configured to transmit the low bandwidth message over low bandwidth communication links to the management controller.