Dynamic deployment of executable recognition resources in distributed camera devices

• US 10,165,180 B2
• Filed: 08/26/2016
• Issued: 12/25/2018
• Est. Priority Date: 08/26/2016
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Liu, Hao, Tang, Dawei, Du, Yuliang
• Primary Examiner(s): Henn, Timothy J
• Application Number:
  US15/248,088
• Publication Number:
  US 20180063422A1
• Time in Patent Office: 851 Days
• CPC Class Codes: H04N 23/617 Upgrading or updating of programs or applications for camera control
• CPC Class Codes: H04N 23/662 by using master/slave camera arrangements for affecting the control of camera image capture, e.g. placing the camera in a desirable condition to capture a desired image
• CPC Class Codes: H04N 23/90 Arrangement of cameras or camera modules, e.g. multiple cameras in TV studios or sports stadiums

Abstract

In one embodiment, a method comprises: identifying a deployment context for execution within one or more distributed camera devices in a distributed camera system, the deployment context including video recognition requirements relative to available capacity in the one or more distributed camera devices; determining an optimized executable recognition resource for the deployment context from available executable recognition resources; and sending, to the one or more distributed camera devices, an instruction for deployment and execution of the optimized executable recognition resource for optimized recognition according to the deployment context.

Claims

1. A method comprising:

determining, by an apparatus in a data network, a deployment context for execution within one or more distributed camera devices in a distributed camera system, including determining, for the deployment context, available capacity in the one or more distributed camera devices, and determining video recognition requirements for the deployment context, the deployment context based on the video recognition requirements relative to the available capacity;

determining an optimized executable video recognition resource for the deployment context from available executable recognition resources, including determining the optimized executable video recognition resource as at least a part of a best match for the video recognition requirements within the available capacity; and

sending, to the one or more distributed camera devices via the data network, an instruction for deployment and execution of the optimized executable video recognition resource for optimized recognition according to the deployment context.

2. The method of claim 1, further comprising:

identifying a second deployment context, distinct from the deployment context, for execution within at least the one or more distributed camera devices;

determining a second optimized executable recognition resource for the second deployment context; and

sending, to the one or more distributed camera devices, a second instruction for deployment and execution of the second optimized executable recognition resource for optimized recognition according to the second deployment context, distinct from the deployment context.

3. The method of claim 2, wherein the second instruction enables the one or more distributed camera devices to switch from the optimized recognition according to the deployment context to the optimized recognition according to the second deployment context.

4. The method of claim 2, wherein the identifying the second deployment context is based on at least one of:

a user request for executing the optimized recognition for an identified event;

predicting the identified event;

or

detection of the identified event by a sensor device.

5. The method of claim 1, wherein the determining of the deployment context includes:

determining the video recognition requirements for an identified event detectable by the one or more one or more distributed camera devices; and

determining the available capacity based on at least one of sensing capabilities or allocated resources in the one or more distributed camera devices.

6. The method of claim 5, wherein:

the determining the optimized executable video recognition resource includes determining a sensor sharing policy enabling a first of the one or more distributed camera devices to share sensor information with a second of the distributed camera devices, based on the first and second distributed camera devices having different sensing capabilities;

the sending including sending, to the first and second distributed camera devices, the sensor sharing policy for the optimized recognition based on the first and second distributed camera devices sharing the sensor information.

7. The method of claim 1, wherein:

the determining of the optimized executable video recognition resource includes determining a feature sharing policy enabling a first of the one or more distributed camera devices to share a recognition feature with a second of the distributed camera devices, the recognition feature determined in response to the first distributed camera device executing the optimized executable video recognition resource on an image captured by the first distributed camera device;

the sending including sending, to the first and second distributed camera devices, the feature sharing policy enabling the first distributed camera device to share the recognition feature with the second distributed camera device.

8. The method of claim 7, wherein:

the instruction is sent to the first distributed camera device;

the determining of the optimized executable video recognition resource includes determining a second optimized executable recognition resource for execution by the second distributed camera device;

the sending includes sending a second instruction, to the second distributed camera device, for deployment and execution of the second optimized executable resource by the second distributed camera device for the optimized recognition according to the deployment context based on the feature sharing policy.

9. An apparatus comprising:

a processor circuit configured for determining a deployment context for execution within one or more distributed camera devices in a distributed camera system, including determining, for the deployment context, available capacity in the one or more distributed camera devices, and determining video recognition requirements for the deployment context, the deployment context based on the video recognition requirements relative to the available capacity, the processor circuit further configured for determining an optimized executable video recognition resource for the deployment context from available executable recognition resources, including determining the optimized executable video recognition resource as at least a part of a best match for the video recognition requirements within the available capacity; and

a device interface circuit configured for sending, to the one or more distributed camera devices via a data network, an instruction for deployment and execution of the optimized executable video recognition resource for optimized recognition according to the deployment context.

10. The apparatus of claim 9, wherein the processor circuit is configured for:

identifying a second deployment context, distinct from the deployment context, for execution within at least the one or more distributed camera devices;

determining a second optimized executable recognition resource for the second deployment context; and

causing the device interface circuit to send, to the one or more distributed camera devices, a second instruction for deployment and execution of the second optimized executable recognition resource for optimized recognition according to the second deployment context, distinct from the deployment context.

11. The apparatus of claim 10, wherein the second instruction enables the one or more distributed camera devices to switch from the optimized recognition according to the deployment context to the optimized recognition according to the second deployment context.

12. The apparatus of claim 10, wherein the processor circuit is configured for identifying the second deployment context based on at least one of:

a user request for executing the optimized recognition for an identified event;

the processor circuit obtaining a prediction predicting the identified event;

or

the processor circuit receiving a message indicating detection of the identified event by a sensor device.

13. The apparatus of claim 9, wherein the processor circuit is configured for:

determining the video recognition requirements for an identified event detectable by the one or more one or more distributed camera devices; and

determining the available capacity based on at least one of sensing capabilities or allocated resources in the one or more distributed camera devices.

14. The apparatus of claim 13, wherein:

the processor circuit is configured for determining a sensor sharing policy enabling a first of the one or more distributed camera devices to share sensor information with a second of the distributed camera devices, based on the first and second distributed camera devices having different sensing capabilities;

the processor circuit configured for causing the device interface circuit to send, to the first and second distributed camera devices, the sensor sharing policy for the optimized recognition based on the first and second distributed camera devices sharing the sensor information.

15. The apparatus of claim 9, wherein:

the processor circuit is configured for determining a feature sharing policy enabling a first of the one or more distributed camera devices to share a recognition feature with a second of the distributed camera devices, the recognition feature determined in response to the first distributed camera device executing the optimized executable video recognition resource on an image captured by the first distributed camera device;

the processor circuit is configured for causing the device interface circuit to send, to the first and second distributed camera devices, the feature sharing policy enabling the first distributed camera device to share the recognition feature with the second distributed camera device.

16. The apparatus of claim 15, wherein:

the instruction is sent to the first distributed camera device;

the processor circuit configured for determining a second optimized executable recognition resource for execution by the second distributed camera device;

the processor circuit is configured for causing the device interface circuit to send a second instruction, to the second distributed camera device, for deployment and execution of the second optimized executable resource by the second distributed camera device for the optimized recognition according to the deployment context based on the feature sharing policy.

17. One or more non-transitory tangible media encoded with logic for execution by a machine and when executed by the machine operable for:

determining, by the machine in a data network, a deployment context for execution within one or more distributed camera devices in a distributed camera system, including determining, for the deployment context, available capacity in the one or more distributed camera devices, and determining video recognition requirements for the deployment context, the deployment context based on the video recognition requirements relative to the available capacity;

determining an optimized executable video recognition resource for the deployment context from available executable recognition resources, including determining the optimized executable video recognition resource as at least a part of a best match for the video recognition requirements within the available capacity; and

sending, to the one or more distributed camera devices via the data network, an instruction for deployment and execution of the optimized executable video recognition resource for optimized recognition according to the deployment context.

18. The one or more non-transitory tangible media of claim 17, further operable for:

identifying a second deployment context, distinct from the deployment context, for execution within at least the one or more distributed camera devices;

determining a second optimized executable recognition resource for the second deployment context; and

sending, to the one or more distributed camera devices, a second instruction for deployment and execution of the second optimized executable recognition resource for optimized recognition according to the second deployment context, distinct from the deployment context.

19. The one or more non-transitory tangible media of claim 17, wherein the determining of the deployment context includes:

determining the video recognition requirements for an identified event detectable by the one or more one or more distributed camera devices; and

determining the available capacity based on at least one of sensing capabilities or allocated resources in the one or more distributed camera devices.

20. The one or more non-transitory tangible media of claim 17, wherein:

the determining of the optimized executable video recognition resource includes determining a feature sharing policy enabling a first of the one or more distributed camera devices to share a recognition feature with a second of the distributed camera devices, the recognition feature determined in response to the first distributed camera device executing the optimized executable video recognition resource on an image captured by the first distributed camera device;

the sending including sending, to the first and second distributed camera devices, the feature sharing policy enabling the first distributed camera device to share the recognition feature with the second distributed camera device.