Scanning camera-based video surveillance system
View Patent ↗A video surveillance system may include at least one sensing unit capable of being operated in a scanning mode and a video processing unit coupled to the sensing unit, the video processing unit to receive and process image data from the sensing unit and to detect scene events and target activity.
1. A video surveillance system comprising:
a processor;
a storage device storing software instructions; and
a camera configured to pan, tilt, and zoom,
wherein the software instructions, when executed by the processor, control the video surveillance system to perform operations comprising:
controlling the camera to scan a scene using a first field of view;
determining a first set of video primitives using first image data received from the camera during the scan, the first set of video primitives describing targets in the first field of view, motion of the camera during the scan, and the scene in the first field of view;
detecting a target of interest in the first field of view using the first set of video primitives;
based on information about the target of interest in the first set of video primitives, controlling the camera to continuously track the target of interest using a second field of view by moving the camera;
determining a second set of video primitives using second image data received from the camera during the continuous tracking, the second set of video primitives describing the target of interest in the second field of view, motion of the camera during the continuous tracking, and the scene in the second field of view;
determining a third field of view of the target of interest using information describing the target of interest in the second set of video primitives and based on image quality criteria;
controlling the camera to obtain the third field of view;
determining that the camera has obtained in the third field of view;
determining a third set of video primitives using third image data received from the camera from the third field of view, the third set of video primitives describing the target of interest in the third field of view;
determining an identity of the target of interest using the third set of video primitives; and
issuing an alert including the identity of the target of interest.
2. The video surveillance system of claim 1 , wherein the operations further comprise:
determining that the camera has failed to obtain the third field of view; and
after failing to obtain the third field of view, controlling the camera to rescan the scene using the first field of view.
3. The video surveillance system of claim 1 , wherein the information about the target of interest in the first set of video primitives comprises an event in the scene or an activity of the target of interest.
4. The video surveillance system of claim 1 , wherein detecting the target of interest in the first field of view comprises:
detecting an event or an activity of the target of interest in the first field of view of the scene by comparing the first set of video primitives to at least one event rule or at least one target activity rule; and
issuing a second alert based on the event or the activity of the target of interest in the first set of video primitives.
5. The video surveillance system of claim 1 , wherein the operations further comprise:
detecting an event or an activity of the target of interest in the second field of view of the scene by comparing the second set of video primitives to at least one scene event rule or at least one target activity rule; and
issuing a third alert based on the event or the activity of the target of interest in the second set of video primitives.
6. The video surveillance system of claim 1 , wherein the operations further comprise:
determining a model of the target of interest based on the second image data, the model of the target of the target of interest describing an appearance of the target of interest and motion of the target of interest.
7. The video surveillance system of claim 1 , wherein the operations further comprise:
periodically switching the camera between recording the second field of view and recording the third field of view.
8. The video surveillance system of claim 1 , wherein the image quality criteria comprises:
an angle of view of a detected event or target activity, and
an image resolution of a detected event or target activity.
9. A computer-implemented method for video surveillance using a camera, the method comprising:
controlling, by a processor, the camera to scan a scene using a first field of view;
determining, by the processor, a first set of video primitives using first image data received from the camera during the scan, the first set of video primitives describing targets in the first field of view, motion of the camera during the scan, and the scene in the first field of view;
detecting, by the processor, a target of interest in the first field of view using the first set of video primitives;
based on information about the target of interest in the first set of video primitives, controlling, by the processor, the camera to continuously track the target of interest using a second field of view by moving the camera;
determining, by the processor, a second set of video primitives using second image data received from the camera during the continuous tracking, the second set of video primitives describing the target of interest in the second field of view, motion of the camera during the continuous tracking, and the scene in the second field of view;
determining, by the processor, a third field of view of the target of interest using information describing the target of interest in the second set of video primitives and based on image quality criteria;
controlling, by the processor, the camera to obtain the third field of view;
determining, by the processor, that the camera has obtained in the third field of view;
determining, by the processor, a third set of video primitives using third image data received from the camera from the third field of view, the third set of video primitives describing the target of interest in the third field of view;
determining, by the processor, an identity of the target of interest using the third set of video primitives; and
issuing, by the processor, an alert including the identity of the target of interest.
10. The method of claim 9 , wherein the information about the target of interest in the first set of video primitives comprises an event in the scene or an activity of the target of interest.
11. The method of claim 9 , wherein detecting the target of interest in the first field of view comprises:
detecting an event or an activity of the target of interest in the first field of view of the scene by comparing the first set of video primitives to at least one scene event rule or at least one target activity rule; and
issuing a second alert based on the event or the activity of the target of interest in the first set of video primitives.
12. The method of claim 9 , wherein the operations further comprise:
detecting an event or an activity of the target of interest in the second field of view of the scene by comparing the second set of video primitives to at least one scene event rule or at least one target activity rule; and
issuing a third alert based on the event or the activity of the target of interest in the second set of video primitives.
13. The method of claim 9 , wherein the operations further comprise:
determining a model of the target of interest based on the second image data, the model of the target of the target of interest describing an appearance of the target of interest and motion of the target of interest.
14. The method of claim 9 , wherein the operations further comprise:
periodically switching the camera between recording the second field of view and recording the third field of view.
15. The method of claim 9 , wherein the image quality criteria comprises:
an angle of view of a detected event or target activity, and
an image resolution of a detected event or target activity.
16. A computer-readable storage device storing software instructions that, when executed by a processor, cause a video surveillance system to perform operations comprising:
controlling the camera to scan a scene using a first field of view;
determining a first set of video primitives using first image data received from the camera during the scan, the first set of video primitives describing targets in the first field of view, motion of the camera during the scan, and the scene in the first field of view;
detecting a target of interest in the first field of view using the first set of video primitives;
based on information about the target of interest in the first set of video primitives, controlling the camera to continuously track the target of interest using a second field of view by moving the camera;
determining a second set of video primitives using second image data received from the camera during the continuous tracking, the second set of video primitives describing the target of interest in the second field of view, motion of the camera during the continuous tracking, and the scene in the second field of view;
determining a third field of view of the target of interest using information about the target of interest in the second set of video primitives and based on image quality criteria;
controlling the camera to obtain the third field of view of the target of interest;
determining that the camera has obtained in the third field of view;
determining a third set of video primitives using third image data received from the camera from the third field of view, the third set of video primitives describing the target in the third field of view;
determining an identity of the target of interest using the third set of video primitives; and
issuing an alert including the identity of the target of interest.