×

Video processing technique for 3D target location identification

  • US 10,089,751 B1
  • Filed: 05/31/2017
  • Issued: 10/02/2018
  • Est. Priority Date: 05/31/2017
  • Status: Active Grant
First Claim
Patent Images

1. A method of determining the location of objects, the method comprising:

  • receiving a video signal from each of a plurality of full motion video cameras observing an area;

    identifying predetermined traits in frames of each video signal, the predetermined traits being associated with the objects;

    demarcating a portion of the video frame with a two-dimensional shape, the demarcated portion containing the identified predetermined traits in the frames of each video signal;

    assigning geospatial information to the two-dimensional shape containing the identified predetermined traits, the geospatial information representing a location of a camera that acquired the frame;

    defining a plurality of three-dimensional shapes, each three-dimensional shape extending from a distance beyond the objects to the location of the camera for each frame of the video signal including the predetermined traits, each three-dimensional shape including a two-dimensional shape with the geospatial information;

    identifying each of the full motion video cameras associated with overlapping coverage of the area that provided the video signal where the two-dimensional shape demarcates the identified predetermined traits;

    calculating intersecting portions of the plural three-dimensional shapes by applying an intersect function to the defined three-dimensional shapes obtained from the full motion video cameras with overlapping coverage of the area, the intersect function outputting an intersecting shape; and

    determining a center of the intersecting shape, the center of the intercepting shape being a determined location of the object.

View all claims
  • 1 Assignment
Timeline View
Assignment View
    ×
    ×