Stream object tracking with delayed object detection
Technology for performing object tracking of an object in a video image using both object detection algorithm and an object tracking algorithm. Because object detection algorithms are generally slower, but accurate, than object tracking algorithms, only certain sampled video frames are subject to the object detection algorithm. The non-sampled frames locate the object using the object tracking algorithm, using the most recent previous object detected frame as a starting frame. In this way, the object detection algorithm is used to periodically, effectively calibrate the object tracking algorithm by using a lagging detection result to hide latency introduced by the object detection algorithm.
1. A computer-implemented method (CIM) comprising:
receiving an ordered and time sequential plurality of video frames from a video image that includes a moving object;
performing object detection on the plurality of video frames to determine an identification of the moving object;
during the performance of object detection, performing object tracking calibration on the plurality of video frames with respect to the moving object using bounding boxes and key frames present in the plurality of video frames; and
during the performance of object detection and subsequent to the performance of object tracking calibration, performing object tracking to locate the moving object in at least some of the video frames to obtain object tracking locations for the moving object;
wherein the performance of object tracking calibration and object tracking during a period of object detection results in a lagging detection result hiding a detection latency associated with the performance of object detection on the plurality of video frames.
2. The CIM of claim 1 further comprising:
performing artificial intelligence vision with accelerated object detection speed due to the hiding of the detection latency by use of the lagging detection result.
3. The CIM of claim 1 wherein:
the plurality of video games represents a sports contest; and
the moving object is player who moves about during the sports contest.
4. The CIM of claim 1 wherein:
the performance of object tracking calibration includes determining locations of the moving object by detecting the bounding boxes.
5. The CIM of claim 1 further comprising:
determining key frames of the plurality of video frames with the key frames being frames where a location change in a location of the moving object is relatively large;
wherein the performance of object detection includes applying an object detection algorithm on at least some of the video frames of the plurality of video frames to locate the moving object in within the video frames.
6. The CIM of claim 1 wherein a sampled frame object location, used as an object tracking location for a given video frame, is the lagging detection result.
7. A computer program product (CPP) comprising:
a set of storage device(s); and
computer code stored collectively in the set of storage device(s), with the computer code including data and instructions to cause a processor(s) set to perform at least the following operations:
receiving an ordered and time sequential plurality of video frames from a video image that includes a moving object,
performing object detection on the plurality of video frames to determine an identification of the moving object,
during the performance of object detection, performing object tracking calibration on the plurality of video frames with respect to the moving object using bounding boxes and key frames present in the plurality of video frames, and
during the performance of object detection and subsequent to the performance of object tracking calibration, performing object tracking to locate the moving object in at least some of the video frames to obtain object tracking locations for the moving object,
wherein the performance of object tracking calibration and object tracking during a period of object detection results in a lagging detection result hiding a detection latency associated with the performance of object detection on the plurality of video frames.
8. The CPP of claim 7 wherein the computer code further includes data and instructions for performing the following operation(s):
performing artificial intelligence vision with accelerated object detection speed due to the hiding of the detection latency by use of the lagging detection result.
9. The CPP of claim 7 wherein:
the plurality of video games represents a sports contest; and
the moving object is player who moves about during the sports contest.
10. The CPP of claim 7 wherein:
the performance of object tracking calibration includes determining locations of the moving object by detecting the bounding boxes.
11. The CPP of claim 7 wherein the computer code further includes data and instructions for performing the following operation(s):
determining key frames of the plurality of video frames with the key frames being frames where a location change in a location of the moving object is relatively large;
wherein the performance of object detection includes applying an object detection algorithm on at least some of the video frames of the plurality of video frames to locate the moving object within the video frames.
12. The CPP of claim 7 wherein a sampled frame object location, used as an object tracking location for a given video frame, is the lagging detection result.
13. A computer system (CS) comprising:
a processor(s) set;
a set of storage device(s); and
computer code stored collectively in the set of storage device(s), with the computer code including data and instructions to cause the processor(s) set to perform at least the following operations:
receiving an ordered and time sequential plurality of video frames from a video image that includes a moving object,
performing object detection on the plurality of video frames to determine an identification of the moving object,
during the performance of object detection, performing object tracking calibration on the plurality of video frames with respect to the moving object using bounding boxes and key frames present in the plurality of video frames, and
during the performance of object detection and subsequent to the performance of object tracking calibration, performing object tracking to locate the moving object in at least some of the video frames to obtain object tracking locations for the moving object,
wherein the performance of object tracking calibration and object tracking during a period of object detection results in a lagging detection result hiding a detection latency associated with the performance of object detection on the plurality of video frames.
14. The CS of claim 13 wherein the computer code further includes data and instructions for performing the following operation(s):
performing artificial intelligence vision with accelerated object detection speed due to the hiding of the detection latency by use of the lagging detection result.
15. The CS of claim 13 wherein:
the plurality of video games represents a sports contest; and
the moving object is player who moves about during the sports contest.
16. The CS of claim 13 wherein:
the performance of object tracking calibration includes determining locations of the moving object by detecting the bounding boxes.
17. The CS of claim 13 wherein the computer code further includes data and instructions for performing the following operation(s):
determining key frames of the plurality of video frames with the key frames being frames where a location change in a location of the moving object is relatively large;
wherein the performance of object detection includes applying an object detection algorithm on at least some of the video frames of the plurality of video frames to locate the moving object within the video frames.
18. The CS of claim 13 wherein a sampled frame object location, used as an object tracking location for a given video frame, is the lagging detection result.