Object detection metadata
A perimeter around a detected object in a frame of image data can be generated in a first coordinate system. The perimeter can be converted from the first coordinate system into a second coordinate system having the same aspect ratio as the first coordinate system. A first metadata entry can include dimensions of image data in the second coordinate system. A second metadata entry can provide a location and dimensions of the converted perimeter in the second coordinate space. Additional metadata can indicate matching objects between frames, position of an object relative to other objects in a frame, a probability that an object is correctly detected, and a total number of objects detected across multiple frames of image data.
1. A method for generating face detection metadata, the method comprising:
generating a perimeter around a detected face in a frame of video data, the perimeter using a first coordinate system;
converting, by a data processing system, the perimeter from the first coordinate system into a second coordinate system, wherein an aspect ratio of dimensions of the perimeter in the first coordinate system matches an aspect ratio of the dimensions in the second coordinate system;
generating a first metadata entry including dimensions of the frame of video data in the second coordinate system;
generating a second metadata entry including information that provides a location and dimensions of the converted perimeter in the second coordinate system.
2. The method of claim 1 , wherein the location comprises a point indicating the top-left corner of a rectangle and the dimensions of the converted perimeter comprise a width and a height of the rectangle.
3. The method of claim 1 , wherein:
the first metadata entry further includes an indication of a total number of detected faces; and
the dimensions of the frame of video data in the second coordinate system comprise a width and a height.
4. The method of claim 1 , further comprising storing the first metadata entry and the second metadata entry in a metadata track corresponding to the video data.
5. The method of claim 4 , wherein the metadata track is configured for time-based metadata storage, the metadata entries in the metadata track being temporally-aligned with the video data.
6. A non-transitory machine readable storage medium storing instructions which when executed by a processor cause a device to perform operations, the operations comprising:
generating a perimeter around a detected object in a frame of video data, the perimeter using a first coordinate system, and the first coordinate system having a first aspect ratio;
converting the perimeter from the first coordinate system into a second coordinate system, wherein dimensions of the second coordinate system have a second aspect ratio matching the first aspect ratio;
generating a first metadata entry including dimensions of the second coordinate system;
generating a second metadata entry including location and dimensions of the converted perimeter in the second coordinate system.
7. The non-transitory machine readable storage medium of claim 6 , wherein the detected object is one of a human face and an animal face.
8. The non-transitory machine readable storage medium of claim 6 , wherein the detected object is physical object.