Back to results
US 8,711,220 B2Granted
Automatic detection of image degradation in enhanced vision systems
AssigneeAireyes, Inc.
InventorCarlo L. M. Tiana
Life of this patent
Estimating the term…
Abstract
Systems and methods are described that directly detect whether or not enhanced vision system images obtained from a moving vehicle are degraded. If they are, image display is turned off to reduce operator distraction. When image quality is restored, image display is turned back on so operators can take advantage of EVS capabilities.
Claims22
A method for detecting enhanced vision system image degradation comprising:
- providing a live video stream from an enhanced vision sensor mounted to a vehicle, the video having a frame rate between one and one hundred frames per second;
- selecting an analysis area in each video frame;
- monitoring a statistical parameter of the video stream as a function of time in the analysis area;
- calculating a long-term average value for the statistical parameter over a period of minutes;
- establishing first and second threshold values for the parameter;
- and,
- declaring the video stream degraded during any time period that:
- starts when the parameter deviates from the long-term average and crosses the first threshold, and ends when the parameter crosses the second threshold.
5 more elements
A system for detecting enhanced vision system image degradation comprising:
- an enhanced vision system sensor mounted to a vehicle, the sensor providing a live video stream having a frame rate between one and one hundred frames per second;
- an image degradation detector that:
- selects an analysis area in each video frame,
- monitors a statistical parameter of the video stream as a function of time in the analysis area,
- calculates a long-term average value for the statistical parameter over a period of minutes, and
- establishes first and second threshold values for the parameter;
- and,
- a display that shows the video stream except during any time period that:
- starts when the parameter deviates from the long-term average and crosses the first threshold, and ends when the parameter crosses the second threshold.
6 more elements
Dependencies are inferred from the claim text.