Systems and methods for tracking objects
Example implementations include a method, apparatus and computer-readable medium in a video surveillance system for tracking an of-interest (OI) object, comprising receiving, from a user interface, a request to track an object in a first video stream from a plurality of video streams captured from a first camera of a plurality of cameras installed in an environment. The implementations further include extracting at least one image of the object from the first video stream. Additionally, the implementations further include generating a unique biometric identifier of the object based on the at least one image. Additionally, the implementations further include detecting, using the unique biometric identifier, the object in a second video stream captured from a second camera of the plurality of cameras, and outputting, on the user interface, the second video stream in response to detecting the object in the second video stream.
1. An apparatus in a video surveillance system for tracking an of-interest (OI) object, comprising:
a memory; and
a processor coupled with the memory and configured to:
receive, from a user interface, a request to track an object in a first video stream from a plurality of video streams, wherein the first video stream is captured from a first camera of a plurality of cameras installed in an environment;
extract at least one image of the object from the first video stream;
generate a unique biometric identifier of the object based on the at least one image;
monitor the plurality of video streams for the object by comparing a candidate unique biometric identifier of a candidate object in at least one of the plurality of video streams with the unique biometric identifier of the object;
detect, based on the monitoring, the object in a second video stream captured from a second camera of the plurality of cameras, wherein the user interface is configured to simultaneously output a subset of the plurality of video streams, the subset comprising the first video stream and the second video stream, wherein a first display area of the first video stream is larger than a second display area of the second video stream; and
increase the second display area of the second video stream to be larger than the first display area of the first video stream in response to detecting the object in the second video stream.
2. The apparatus of claim 1 , wherein to receive, from the user interface, the request to track the object in the first video stream from the plurality of video streams the processor is further configured to:
receive a selection on a frame of the first video stream, wherein the selection comprises a visual boundary encompassing the object; and
receive an enroll request indicating to enroll the object within the visual boundary for automatic surveillance tracking.
3. The apparatus of claim 1 , wherein to output the second video stream comprises replacing the first video stream with the second video stream in a viewing window of the user interface.
4. The apparatus of claim 1 , wherein to generate the unique biometric identifier the processor is further configured to:
isolate an object image portion from a background image portion in the at least one image; and
generate the unique biometric identifier based on the object image portion in the at least one image.
5. The apparatus of claim 1 , wherein the unique biometric identifier is an input feature vector of a machine learning algorithm configured to detect objects in an image that match the input feature vector.
6. The apparatus of claim 5 , wherein to detect the object in the second video stream comprises to:
detect the candidate object in the second video stream;
extract at least one candidate image of the candidate object from the second video stream;
generate the candidate unique biometric identifier of the candidate object based on the at least one candidate image; and
wherein to detect, based on the monitoring, the object in the second video stream is in response to determining, via the machine learning algorithm, that the candidate unique biometric identifier of the candidate object matches the unique biometric identifier of the object.
7. The apparatus of claim 1 , wherein the unique biometric identifier is a floating point number.
8. The apparatus of claim 1 , wherein to extract the at least one image of the object from the first video stream the processor is further configured to extract the at least one image from all frames in the first video stream in which the object appears during a period of time from receiving the request.
9. The apparatus of claim 1 , wherein the object is a person, and wherein the unique biometric identifier is based on non-facial features of the person.
10. The apparatus of claim 9 , wherein the processor is further configured to:
extract at least one additional image depicting a face of the person from the first video stream;
generate a unique facial biometric identifier of the face of the person based on the at least one additional image; and
wherein to monitor the plurality of video streams for the object the processor is further configured to compare a candidate unique facial biometric identifier of the candidate object with the unique facial biometric identifier.
11. The apparatus of claim 1 , wherein to extract the at least one image the processor is further configured to:
receive a selection on a frame of the first video stream, wherein the selection comprises a visual boundary encompassing the object; and
identify the selection as the at least one image for extraction.
12. The apparatus of claim 1 , wherein to output the second video stream comprises to:
set the first display area of the first video stream and the second display area of the second video stream to be larger than remaining video streams in the subset in response to determine that the object appears in the first video stream and the second video stream simultaneously.
13. The apparatus of claim 12 , wherein the first display area of the first video stream and the second display area of the second video stream are equal.
14. A method in a video surveillance system for tracking an of-interest (OI) object, comprising:
receiving, from a user interface, a request to track an object in a first video stream from a plurality of video streams, wherein the first video stream is captured from a first camera of a plurality of cameras installed in an environment;
extracting at least one image of the object from the first video stream;
generating a unique biometric identifier of the object based on the at least one image;
monitoring the plurality of video streams for the object by comparing a candidate unique biometric identifier of a candidate object in at least one of the plurality of video streams with the unique biometric identifier of the object;
detecting, based on the monitoring, the object in a second video stream captured from a second camera of the plurality of cameras, wherein the user interface is configured to simultaneously output a subset of the plurality of video streams, the subset comprising the first video stream and the second video stream, wherein a first display area of the first video stream is larger than a second display area of the second video stream; and
increasing the second display area of the second video stream to be larger than the first display area of the first video stream in response to detecting the object in the second video stream.
15. The method of claim 14 , wherein receiving, from the user interface, the request to track the object in the first video stream from the plurality of video streams further comprises:
receiving a selection on a frame of the first video stream, wherein the selection comprises a visual boundary encompassing the object; and
receiving an enroll request indicating to enroll the object within the visual boundary for automatic surveillance tracking.
16. The method of claim 14 , wherein outputting the second video stream comprises replacing the first video stream with the second video stream in a viewing window of the user interface.
17. The method of claim 14 , wherein generating the unique biometric identifier further comprises:
isolating an object image portion from a background image portion in the at least one image; and
generating the unique biometric identifier based on the object image portion in the at least one image.
18. The method of claim 14 , wherein the unique biometric identifier is an input feature vector of a machine learning algorithm configured to detect objects in an image that match the input feature vector.
19. The method of claim 18 , wherein detecting the object in the second video stream comprises:
detecting the candidate object in the second video stream;
extracting at least one candidate image of the candidate object from the second video stream;
generating the candidate unique biometric identifier of the candidate object based on the at least one candidate image; and
wherein detecting, based on the monitoring, the object in the second video stream is in response to determining, via the machine learning algorithm, that the candidate unique biometric identifier of the candidate object matches the unique biometric identifier of the object.
20. The method of claim 14 , wherein the unique biometric identifier is a floating point number.
21. The method of claim 14 , wherein extracting the at least one image of the object from the first video stream further comprises extracting the at least one image from all frames in the first video stream in which the object appears during a period of time from receiving the request.
22. The method of claim 14 , wherein the object is a person, and wherein the unique biometric identifier is based on non-facial features of the person.
23. The method of claim 22 , further comprising:
extracting at least one additional image depicting a face of the person from the first video stream;
generating a unique facial biometric identifier of the face of the person based on the at least one additional image; and
wherein monitoring the plurality of video streams for the object further comprises comparing a candidate unique facial biometric identifier of the candidate object with the unique facial biometric identifier.
24. The method of claim 14 , wherein extracting the at least one image further comprises:
receiving a selection on a frame of the first video stream, wherein the selection comprises a visual boundary encompassing the object; and
identifying the selection as the at least one image for extraction.
25. The method of claim 14 , wherein outputting the second video stream comprises:
setting the first display area of the first video stream and the second display area of the second video stream to be larger than remaining video streams in the subset in response to determining that the object appears in the first video stream and the second video stream simultaneously.
26. The method of claim 25 , wherein the first display area of the first video stream and the second display area of the second video stream are equal.