Stationary target detection by exploiting changes in background model

Patent metadata

Field	Value
Publication number	US-11170225-B2
Application number	US-202016777203-A
Country	US
Kind code	B2
Filing date	Jan 30, 2020
Priority date	Sep 4, 2007
Publication date	Nov 9, 2021
Grant date	Nov 9, 2021

Abstract

A sequence of video frames of an area of interest is obtained. A first background model of the area of interest is constructed based on a first parameter. A second background model of the area of interest is constructed based on a second parameter, the second parameter being different from the first parameter. A difference between the first and second background models is determined. A stationary target is determined based on the determined difference. An alert concerning the stationary target is generated.

First claim

The invention claimed is:

1. A computer-implemented method for processing one or more video frames, the method comprising:
generating, using the one or more video frames, a first background model with an associated first learning rate;
generating, using the one or more video frames, a second background model with an associated second learning rate different than the first learning rate;
comparing the first background model and the second background model to generate a comparison; and
generating one or more blobs in response to the comparison.

2. The computer-implemented method as claimed in claim 1 further comprising classifying each of the one or more blobs using an artificial neural network classifier.

3. The computer-implemented method as claimed in claim 2 wherein the artificial neural network classifier is configured to classify the one or more blobs in at least a first class of object and a second class of object different than the first class.

4. The computer-implemented method as claimed in claim 3 wherein the first class of object is a person.

5. The computer-implemented method as claimed in claim 4 wherein the first class of object is a vehicle.

6. The computer-implemented method as claimed in claim 4 wherein the first class of object is a piece of luggage.

7. The computer-implemented method as claimed in claim 1 wherein the second learning rate is slower than the first learning rate.

8. The computer-implemented method as claimed in claim 1 further comprising performing salience filtering to filter out erroneous results.

9. The computer-implemented method as claimed in claim 1 further comprising carrying out classifications of the one or more blobs by performing a plurality of gradients calculations taking into account the first and second background models.

10. A non-transitory computer-readable medium containing instructions which, when executed on a processor, perform a method for processing one or more video frames, the method comprising:
generating, using the one or more video frames, a first background model with an associated first learning rate;
generating, using the one or more video frames, a second background model with an associated second learning rate different than the first learning rate;
comparing the first background model and the second background model to generate a comparison; and
generating one or more blobs in response to the comparison.

11. The non-transitory computer-readable medium as claimed in claim 10 wherein the method further includes classifying each of the one or more blobs using an artificial neural network classifier.

12. The non-transitory computer-readable medium as claimed in claim 11 wherein the artificial neural network classifier is configured to classify the one or more blobs in at least a first class of object and a second class of object different than the first class.

13. The non-transitory computer-readable medium as claimed in claim 12 wherein the first class of object is a person.

14. The non-transitory computer-readable medium as claimed in claim 12 wherein the first class of object is a vehicle.

15. The non-transitory computer-readable medium as claimed in claim 12 wherein the first class of object is a piece of luggage.

16. The non-transitory computer-readable medium as claimed in claim 10 wherein the second learning rate is slower than the first learning rate.

17. The non-transitory computer-readable medium as claimed in claim 10 wherein the method further includes performing salience filtering to filter out erroneous results.

18. The non-transitory computer-readable medium as claimed in claim 10 wherein the method further includes carrying out classifications of the one or more blobs by performing a plurality of gradients calculations taking into account the first and second background models.

Assignees

• Avigilon Fortress Corp

Inventors

• Hassan-Shafique Khurram
• Venetianer Peter L
• Yu Li
• Liu Haiying

Classifications

• G08B13/19602 Primary

  Image analysis to detect motion of the intruder, e.g. by frame subtraction · CPC title

• G06V20/52

  Surveillance or monitoring of activities, e.g. for recognising suspicious objects (recognising microscopic objects G06V20/69) · CPC title

• G06F18/24

  Classification techniques · CPC title

• G06K9/00771 Primary

  Physics · mapped topic

• G06K9/6267

  Physics · mapped topic