US 12,251,994 B1
Automotive glare reduction system
Ramesh Balasubramanyan, Salem, OR (US)
Filed by Ramesh Balasubramanyan, Salem, OR (US)
Filed on Jul. 23, 2021, as Appl. No. 17/384,112.
Claims priority of provisional application 63/060,049, filed on Aug. 1, 2020.
Int. Cl. B60J 3/04 (2006.01); G06V 20/58 (2022.01); G06V 40/18 (2022.01)
CPC B60J 3/04 (2013.01) [G06V 20/584 (2022.01); G06V 40/18 (2022.01)] 12 Claims
OG exemplary drawing
 
5. A system for a vehicle for reducing glare experienced by a driver of the vehicle comprising:
a camera mounted on the vehicle, facing forward, configured to transmit a stream of images;
a display screen mounted on a windshield of the vehicle, the display screen divided into a plurality of pixels, each pixel individually controllable for at least two transparency states, including a high transparency state and a low transparency state, wherein each pixel in the high transparency state allows more light to pass through than in the low transparency state;
a processor configured for performing the steps of:
capturing a video frame image from the stream of images,
detecting one or more headlights within the video frame image;
determining current measured positions of the one or more headlights within the video frame image;
determining locations of one or more blots on the display screen, based on the current measured positions of the one or more headlights within the video frame image and a current set of setup information, including setup locations of the one or more blots;
sending commands to put the one or more blots on the display screen at the determined locations by setting pixels within the one or more blots to the low transparency state;
controls configured for sending control signals to the processor for changing aspects of the one or more blots, including setup location;
wherein detecting one or more headlights within the video frame image further comprises:
scanning the video frame image with a first Haar Cascade classifier to filter out lights that are not headlights, determine coordinates within the video frame image of the one or more headlights;
extracting a subset image of the video frame image based on the coordinates of the one or more headlights; and
scanning the subset image with a second Haar Cascade classifier to determine the current measured positions of each of the one or more headlights within the subset image.