US 9,544,489 B2 · Effortless navigation across cameras and cooperative control of cameras

Abstract

This application discloses a system and related methods (called Fast Track) to effortlessly navigate across multiple cameras, e.g. video surveillance cameras distributed in a facility. Fast Track simplifies the act of following a suspicious individual roaming in a facility, e.g. by eliminating the need for security operators to memorize camera placement and coverage, and to manually control pan-tilt-zoom (hereinafter “PTZ”) cameras, and by simplifying the process to reduce the risk of operator error. Combinations of novel scoring and control methods generally provide an automatic identification of optimal cameras and pre-emptive re-direction of PTZ cameras. Extensions to these methods enable two new capabilities: Cooperative PTZ Control (e.g. coordinating PTZs to cover a 3D map viewport optimally) and Self-Healing Perimeters, to automatically reassign and redirect multiple PTZ cameras to fill gaps in a security perimeter.

Claims16

Dependent claims

A method for automatically selecting at least one camera among a plurality of cameras disposed in a 3D map each of the plurality of cameras comprising a field of view and a 3D position, the method comprising:

a) for each of the plurality of cameras, calculating a 3D projective transformation of the 3D map into the field of view of the camera;
b) determining a 3D point anywhere in the 3D map;
c) for each of the plurality of cameras, calculating a per-camera score with respect to the 3D point, each of the per-camera scores being calculated at least with respect to an indication of the visibility of the point in the calculated 3D projective transformation of 3D map into the field of view of the camera, the calculation comprising:
i) extrapolating a plurality of sample 3D points along an extrapolated sampling distribution 3D shape generated around the 3D point;
ii) calculating a per-sample score for each of the sample 3D points, and combining the per-sample scores to produce the per-camera score;
d) automatically selecting the at least one camera with the highest per-camera score;
and
e) performing an action on the selected at least one camera.

5 more elements

A method for cooperative control of a plurality of cameras disposed in a 3D map, each of the plurality of cameras comprising a 3D direction, a zoom, and a 3D position in the map, the method comprising:

a) for each of the plurality of cameras, calculating a 3D projective transformation of the 3D map into the field of view of the camera;
b) distributing weighted sample 3D points on at least one area in the 3D map to be monitored by at least some of the plurality of cameras, wherein each of the sample 3D points is being assigned a weight relative to its importance;
c) determining at least a subset of the plurality of cameras in which each of the cameras of the subset of the plurality of cameras is capable of viewing at least some of the sample 3D points;
d) for each of the cameras part of the subset of the plurality of cameras, calculating a per-camera score with respect to at least some of the sample 3D points, each of the per-camera scores being calculated at least with respect to an indication of the visibility of the at least some of the sample points in the calculated 3D projective transformation of the 3D map into the field of view of the camera;
and
e) changing, if necessary, the direction and/or the zoom and/or the position of each camera of the subset of the plurality of cameras such that each camera of the subset of the plurality of cameras covers at least some of the weighted sample 3D points and such that most of the weighted sample 3D points are covered by at least one of the cameras of the subset of the plurality of cameras.

3 more elements

Dependencies are inferred from the claim text.

Effortless navigation across cameras and cooperative control of cameras

Life of this patent

Effortless navigation across cameras and cooperative control of cameras

Life of this patent

Abstract

Claims16