Back to results
US 2022/0036084 A1Published application
Keyframe Selection for Computer Vision Analysis
AssigneeNCR Corporation
InventorJoshua Migdal
Life of this patent
Estimating the term…
Abstract
Upon a triggering event, frames of a video feed are analyzed for activity or motion. One frame or a subset of the frames are selected as a keyframe/keyframes based on one or measures associated with one or more metrics derived from the frames. The keyframe/keyframes optimally reflects an activity or degree of motion across the frames. The keyframe/keyframes are then provided as input to a computer vision or machine learning application to improve the efficiency, accuracy, and/or precision of results produced by or decisions made by the application.
Claims20
A method, comprising:
- determining a trigger event;
- obtaining frames of a video feed based on the trigger event;
- identifying a termination event from a subset of the frames;
- selecting at least one keyframe as an optimal representation of an activity or a motion within the subset of the frames based on a calculated measure;
- and
- providing the at least one keyframe as the optimal representation of the activity or the motion for the subset of the frames.
3 more elements
A method, comprising:
- selecting a subset of frames from a video feed based on a trigger event;
- calculating, for each of the subset of frames, a calculated measure based on at least one metric;
- determining at least one keyframe from the subset of frames based on the corresponding calculated measure using the corresponding at least one metric;
- and
- providing the at least one keyframe in place of the subset of frames as input to a computer vision application or a machine learning application that monitors an area captured by a camera and represented in the subset of frames.
2 more elements
A system, comprising:
- a camera;
- a device comprising a processor and a non-transitory computer-readable storage medium comprising executable instructions;
- and
- the executable instructions when executed by the processor from the non-transitory computer-readable storage medium causing the processor to perform processing comprising:
- obtaining a subset of frames from a video feed captured by the camera for an area based on at least one trigger event;
- deriving at least one motion metric for each of the subset of frames;
- calculating a measure for each of the subset of frames based on the corresponding at least one motion metric;
- selecting at least one keyframe from the subset of frames based on the corresponding measure associated with each of the subset of frames;
- and
- providing the at least one keyframe as an optimal frame determined from the subset of frames.
7 more elements
Dependencies are inferred from the claim text.