Method and device for associating frames in a video of an activity of a person with an event
First Claim
Patent Images
1. A method comprising:
- recording a video of an activity of a person;
storing a time-series comprising a plurality of sensor data obtained from a sensor assembly, wherein the sensor assembly comprises at least one sensor coupled to the person while the person is performing the activity;
synchronizing the video with the plurality of sensor data;
detecting an event in the time-series by at least;
preprocessing the time-series;
segmenting the time-series into a plurality of windows;
detecting each of the plurality of windows comprising at least one outlier;
extracting a plurality of features from the time-series in each of the plurality of windows; and
estimating an event class for each of the plurality of windows based on the plurality of features extracted from the time-series in each of the plurality of windows; and
associating the event with at least one corresponding frame in the video showing the event.
1 Assignment
0 Petitions
Accused Products
Abstract
Described are methods and systems for associating frames in a video of an activity of a person with an event. The methods include recording a video of an activity of a person; storing a time-series of a plurality of sensor data (82) obtained from a sensor assembly (12) of at least one sensor (31, 32, 33, 34, 35) coupled to the person while the person is performing the activity; synchronizing the video with the sensor data (82); detecting an event in the time-series; and associating the event with at least one corresponding frame in the video showing the event.
23 Citations
36 Claims
-
1. A method comprising:
-
recording a video of an activity of a person; storing a time-series comprising a plurality of sensor data obtained from a sensor assembly, wherein the sensor assembly comprises at least one sensor coupled to the person while the person is performing the activity; synchronizing the video with the plurality of sensor data; detecting an event in the time-series by at least; preprocessing the time-series; segmenting the time-series into a plurality of windows; detecting each of the plurality of windows comprising at least one outlier; extracting a plurality of features from the time-series in each of the plurality of windows; and estimating an event class for each of the plurality of windows based on the plurality of features extracted from the time-series in each of the plurality of windows; and associating the event with at least one corresponding frame in the video showing the event. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27)
-
-
28. A system comprising:
-
at least one camera configured to record a video of an activity of a person; at least one sensor assembly comprising at least one sensor, the sensor assembly configured to be coupled to the person while the person is performing the activity; a non-transitory memory configured to store a time-series comprising a plurality of sensor data obtained from the at least one sensor; and a processor configured to; synchronize the video with the plurality of sensor data; detect an event in the time-series by at least; preprocessing the time-series; segmenting the time-series into a plurality of windows; detecting each of the plurality of windows comprising at least one outlier; extracting a plurality of features from the time-series in each of the plurality of windows; and estimating an event class for each of the plurality of windows based on the plurality of features extracted from the time-series in each of the plurality of windows; and associate the event with at least one corresponding frame in the video showing the event. - View Dependent Claims (29, 30, 31, 32, 33)
-
-
34. A non-transitory computer-readable medium comprising one or more software applications configured to be executed by a processor, the one or more software applications configured to:
-
record a video of an activity of a person; store a time-series comprising a plurality of sensor data obtained from a sensor assembly, wherein the sensor assembly comprises at least one sensor coupled to the person while the person is performing the activity; synchronize the video with the plurality of sensor data; detect an event in the time-series by at least; preprocessing the time-series; segmenting the time-series into a plurality of windows; detecting each of the plurality of windows comprising at least one outlier; extracting a plurality of features from the time-series in each of the plurality of windows; and estimating an event class for each of the plurality of windows based on the plurality of features extracted from the time-series in each of the plurality of window; and associate the event with at least one corresponding frame in the video showing the event. - View Dependent Claims (35, 36)
-
Specification