Object rejection system and method

• US 10,402,643 B2
• Filed: 06/15/2016
• Issued: 09/03/2019
• Est. Priority Date: 06/15/2016
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google LLC (Alphabet Inc.)
• Inventors: James, Bryan, Modi, Yash, Purohit, Aveek Ravishekhar, Herman, Kenneth Louis, Malhotra, Mark Rajan
• Primary Examiner(s): Yenke, Brian P
• Application Number:
  US15/183,471
• Publication Number:
  US 20170364743A1
• Time in Patent Office: 1,175 Days
• CPC Class Codes: G06F 2218/14 by matching peak patterns
• CPC Class Codes: G06V 10/143 Sensing or illuminating at different wavelengths
• CPC Class Codes: G06V 20/52 Surveillance or monitoring of activities, e.g. for recognising suspicious objects recognising microscopic objects G06V20/69

Abstract

A system and method for detecting human intruders while rejecting/ignoring an occupant'"'"'s registered pet. An object detection system is configured to detect an object that is present in a monitored area and generate a signal output relative to the type of object. A signature processor is configured to receive the generated signal output and produce an object signature, and compare a threshold signature to the object signature, wherein the threshold signature is generated using a photograph of a reserved object, and wherein the object detection system rejects the detected object when the object signature is determined to be similar to the threshold signatures.

Claims

1. A system comprising:

an object detection system comprising an infrared sensor configured to detect an object that is present in a monitored area and generate a signal output representative of the type of object, wherein the signal output is based on infrared radiation detected from the object; and

a signature processor configured to;

receive the generated signal output and produce an object signature based on the infrared radiation detected from the object; and

compare a threshold signature to the object signature to determine if the object is a reserved object or an unauthorized object,

wherein the threshold signature is generated using a non-infrared photograph of the reserved object and is generated using previously detected infrared radiation from one or more objects identified as having similar non-infrared radiation characteristics to the reserved object based on the non-infrared photograph of the reserved object, wherein the one or more objects do not include the reserved object, and

wherein the object detection system rejects the detected object when the object signature is determined to be similar to the threshold signature.

2. The system of claim 1, wherein the signature processor compares a peak amplitude of the threshold signature to a peak amplitude of the object signature to determine whether the object is the reserved object or an unauthorized object.

3. The system of claim 2, wherein the non-infrared photograph is a picture of an occupant'"'"'s pet.

4. The system of claim 3, wherein estimated characteristics of a part from the non-infrared photograph are used to generate the threshold signature.

5. The system of claim 2, further comprising a service provider server configured to:

estimate characteristics of the reserved object in the non-infrared photograph;

generate the threshold signature using the estimated characteristics; and

forward to the object detection system the threshold signature.

6. The system of claim 5, wherein the service provider server is further configured to retrieve from a pattern database a pattern signature that is associated with an object having similar characteristics as the estimated characteristics, wherein the retrieved pattern signature is the threshold signature.

7. The system of claim 6, wherein the estimated characteristics include one or more of height, weight, length, type of object, and hair length.

8. The system of claim 6,

wherein the pattern database includes a plurality of pattern signatures, each pattern signature associated with an object having object characteristics, and

wherein the threshold signature is generated based on a selection of the pattern signature having objects characteristics similar to the estimated characteristics.

9. The system of claim 5, wherein the reserved object in the non-infrared photograph is an occupant'"'"'s pet.

10. The system of claim 2, wherein the object detection system comprises:

the infrared sensor configured to detect the presence of the object in the monitored area and produce a sensor signal; and

a sensor processor configured to generate the signal output based on the sensor signal.

11. The system of claim 10, wherein the sensor is a motion sensor.

12. A method comprising:

detecting, by an infrared (PIR) sensor, an object that is present in a monitored area;

generating a signal output representative of the type of object, wherein the signal output is based on infrared radiation detected from the object;

generating an object signature based on the infrared radiation detected from the object using the signal output; and

comparing a threshold signature to the object signature to determine if the object is a reserved object or an unauthorized object,

wherein the threshold signature is generated using a non-infrared photograph of a reserved object and is generated using previously detected infrared radiation from one or more objects identified as having similar non-infrared radiation characteristics to the reserved object based on the non-infrared photograph of the reserved object, wherein the one or more objects do not include the reserved object and

wherein the object detection system rejects the detected object when the object signature is determined to be similar to the threshold signature.

13. The method of claim 12, further comprising comparing a peak amplitude of the threshold signature to a peak amplitude of the object signature to determine whether the object is the reserved object or an unauthorized object.

14. The method of claim 13, wherein the non-infrared photograph is a picture of an occupant'"'"'s pet.

15. The method of claim 14, wherein estimated characteristics of the part from the non-infrared photograph are used to generate the threshold signature.

16. The method of claim 13, further comprising:

estimating characteristics of the reserved object in the non-infrared photograph;

generating the threshold signature using the estimated characteristics; and

forwarding to an object detection system the threshold signature.

17. The method of claim 16, further comprising retrieving from a pattern database a pattern signature that is associated with an object having similar characteristics as the estimated characteristics, wherein the retrieved pattern signature is the threshold signature.

18. The method of claim 17, wherein the estimated characteristics include one or more of height, weight, length, type of object, and hair length.

19. The method of claim 17,

wherein the pattern database includes a plurality of pattern signatures, each pattern signature associated with an object having object characteristics, and

wherein the threshold signature is generated based on a selection of the pattern signature having objects characteristics similar to the estimated characteristics.

20. The method of claim 16, wherein the reserved object in the non-infrared photograph is an occupant'"'"'s pet.