Detecting and sharing events of interest using panoptic computer vision systems
In one example, a method performed by a processing system including at least one processor includes acquiring a plurality of video volumes of an environment from a plurality of cameras, wherein at least two individual video volumes of the plurality of video volumes depict the environment from different viewpoints, generating a panoptic video feed of the environment from the plurality of video volumes, detecting an event of interest occurring in the panoptic video feed, and isolating a video volume of the event of interest to produce a video excerpt.
1. A method comprising:
acquiring, by a processing system including at least one processor, a plurality of video volumes of an environment from a plurality of cameras, wherein at least two individual video volumes of the plurality of video volumes depict the environment from different viewpoints, and wherein the plurality of cameras includes at least one camera mounted to a vehicle and a camera of a user endpoint device of a passenger of the vehicle;
generating, by the processing system in real time as the plurality of video volumes is acquired, a panoptic video feed of the environment from the plurality of video volumes;
detecting, by the processing system, an event of interest occurring in the panoptic video feed, wherein the detecting comprises using a computerized image analysis technique to automatically identify a presence of a subject whose appearance is included in a list of one or more predefined events of interest in real time while the subject is present in the environment; and
isolating, by the processing system, a video volume of the event of interest to produce a video excerpt,
wherein the processing system is operated by an operator of a tour in the environment, wherein the at least one camera mounted to the vehicle is controlled by the operator of the tour and is calibrated to facilitate the generating of the panoptic video feed, and wherein the user endpoint device is controlled by the passenger who is independent of the operator of the tour and the camera of the user endpoint device is not calibrated to facilitate the generating of the panoptic video feed.
2. The method of claim 1 , wherein data from a sensor is used to track a trajectory of the event of interest.
3. The method of claim 1 , wherein the isolating isolates one or more video volumes of the event of interest in both space and time.
4. The method of claim 1 , wherein the isolating utilizes an image processing technique to enhance the video excerpt.
5. The method of claim 1 , further comprising:
delivering, by the processing system, the video excerpt to a display device.
6. The method of claim 5 , wherein the display device comprises a personal user endpoint device of a person who is present in the environment.
7. The method of claim 5 , wherein the display device comprises a personal user endpoint device of a person who is viewing the environment remotely.
8. The method of claim 5 , wherein the display device is embedded in a window of the vehicle to which at least one camera of the plurality of cameras is mounted.
9. The method of claim 5 , wherein the video excerpt is delivered as part of a continuous stream of video data.
10. The method of claim 1 , further comprising:
labeling the event of interest depicted in the video excerpt in response to a signal from a human operator.
11. The method of claim 10 , wherein the signal comprises at least one of: a spoken signal, a text-based signal, or a gesture.
12. The method of claim 1 , wherein the video volume of the event of interest is annotated with data from a data source.
13. A non-transitory computer-readable medium storing instructions which, when executed by a processing system including at least one processor, cause the processing system to perform operations, the operations comprising:
acquiring a plurality of video volumes of an environment from a plurality of cameras, wherein at least two individual video volumes of the plurality of video volumes depict the environment from different viewpoints, and wherein the plurality of cameras includes at least one camera mounted to a vehicle and a camera of a user endpoint device of a passenger of the vehicle;
generating in real time, as the plurality of video volumes is acquired, a panoptic video feed of the environment from the plurality of video volumes;
detecting an event of interest occurring in the panoptic video feed, wherein the detecting comprises using a computerized image analysis technique to automatically identify a presence of a subject whose appearance is included in a list of one or more predefined events of interest in real time while the subject is present in the environment; and
isolating a video volume of the event of interest to produce a video excerpt,
wherein the processing system is operated by an operator of a tour in the environment, wherein the at least one camera mounted to the vehicle is controlled by the operator of the tour and is calibrated to facilitate the generating of the panoptic video feed, and wherein the user endpoint device is controlled by the passenger who is independent of the operator of the tour and the camera of the user endpoint device is not calibrated to facilitate the generating of the panoptic video feed.
14. A system comprising:
a processing system including at least one processor; and,
a non-transitory computer-readable medium storing instructions which, when executed by the processing system, cause the processing system to perform operations, the operations comprising:
acquiring a plurality of video volumes of an environment from a plurality of cameras, wherein at least two individual video volumes of the plurality of video volumes depict the environment from different viewpoints, and wherein the plurality of cameras includes at least one camera mounted to a vehicle and a camera of a user endpoint device of a passenger of the vehicle;
generating in real time, as the plurality of video volumes is acquired, a panoptic video feed of the environment from the plurality of video volumes;
detecting an event of interest occurring in the panoptic video feed, wherein the detecting comprises using a computerized image analysis technique to automatically identify a presence of a subject whose appearance is included in a list of one or more predefined events of interest in real time while the subject is present in the environment; and
isolating a video volume of the event of interest to produce a video excerpt,
wherein the processing system is operated by an operator of a tour in the environment, wherein the at least one camera mounted to the vehicle is controlled by the operator of the tour and is calibrated to facilitate the generating of the panoptic video feed, and wherein the user endpoint device is controlled by the passenger who is independent of the operator of the tour and the camera of the user endpoint device is not calibrated to facilitate the generating of the panoptic video feed.
15. The method of claim 1 , wherein the panoptic video feed presents a 360 degree view of the environment.
16. The method of claim 1 , wherein the vehicle comprises a vehicle that is at least one of: autonomous or semi-autonomous.
17. The method of claim 16 , wherein the processing system sends instructions to a navigation system of the vehicle in response to the detecting to cause the vehicle to follow the event of interest.
18. The non-transitory computer-readable medium of claim 13 , wherein the vehicle comprises a vehicle that is at least one of: autonomous or semi-autonomous.
19. The non-transitory computer-readable medium of claim 18 , wherein the processing system sends instructions to a navigation system of the vehicle in response to the detecting to cause the vehicle to follow the event of interest.
20. The system of claim 14 , wherein the vehicle comprises a vehicle that is at least one of: autonomous or semi-autonomous.