Video processing for embedded information card localization and content extraction
Metadata for one or more highlights of a video stream may be extracted from one or more card images embedded in the video stream. The highlights may be segments of the video stream, such as a broadcast of a sporting event, that are of particular interest. According to one method, video frames of the video stream are stored. One or more information cards embedded in a decoded video frame may be detected by analyzing one or more predetermined video frame regions. Image segmentation, edge detection, and/or closed contour identification may then be performed on identified video frame region(s). Further processing may include obtaining a minimum rectangular perimeter area enclosing all remaining segments, which may then be further processed to determine precise boundaries of information card(s). The card image(s) may be analyzed to obtain metadata, which may be stored in association with at least one of the video frames.
1. A method of extracting a card image from a video frame, the method comprising:
selecting a video frame region that forms a portion of the video frame;
modifying pixel values of a set of pixels adjacent to a boundary of the video frame region;
removing background from the video frame region based on the modified pixel values;
generating an edge image based on the video frame region with removed background;
identifying contours in the edge image; and
extracting the card image as a region enclosed by an area encompassing the contours.
2. The method of claim 1 , wherein the selecting the video frame region comprises cropping the video frame to isolate the video frame region.
3. The method of claim 1 , wherein the selecting the video frame region comprises:
determining whether a location of the card image is known;
based on determining that the location of the card image is known, selecting the portion of the video frame from the known location as the video frame region; and
based on determining that the location of the card image is not known, sequentially processing a plurality of regions of the video frame to identify the video frame region.
4. The method of claim 1 , wherein the modifying the pixel values comprises setting the pixel values to a black level.
5. The method of claim 1 , further comprising:
approximating the identified contours as polygons, wherein the extracting the card image further comprises:
extracting the card image as the region enclosed by the area as a minimum rectangular perimeter encompassing all the polygons.
6. The method of claim 5 , further comprising:
counting modified pixels for at least one boundary of the area; and
based on determining that a number of the modified pixels exceeds a predetermined threshold, moving the at least one boundary inward to generate an adjusted area.
7. The method of claim 6 , further comprising validating the adjusted area by:
counting a first number of pixels in the video frame region;
counting a second number of pixels in the area;
counting a third number of pixels in the adjusted area; and
determining that the adjusted area is valid based on comparing the first number of pixels, the second number of pixels, and the third number of pixels.
8. The method of claim 1 , further comprising:
generating a metadata associated with the video frame based analyzing the extracted card image.
9. The method of claim 8 , wherein the video frame is within a highlight of a video stream of a sporting event broadcast, and wherein the metadata is descriptive of a status of the sporting event during the highlight.
10. The method of claim 9 , further comprising:
presenting, at an output device, the metadata during viewing of the highlight.
11. A system for extracting a card image from a video frame, the system comprising:
a processor; and
a non-transitory storage medium storing computer program instructions, that when executed by the processor cause the system to perform operations comprising:
selecting a video frame region that forms a portion of the video frame;
modifying pixel values of a set of pixels adjacent to a boundary of the video frame region;
removing background from the video frame region based on the modified pixel values;
generating an edge image based on the video frame region with removed background;
identifying contours in the edge image; and
extracting the card image as a region enclosed by an area encompassing the contours.
12. The system of claim 11 , wherein the operation of selecting the video frame region comprises cropping the video frame to isolate the video frame region.
13. The system of claim 11 , wherein the operation of selecting the video frame region comprises:
determining whether a location of the card image is known;
based on determining that the location of the card image is known, selecting the portion of the video frame from the known location as the video frame region; and
based on determining that the location of the card image is not known, sequentially processing a plurality of regions of the video frame to identify the video frame region.
14. The system of claim 11 , wherein the operation of modifying the pixel values comprises setting the pixel values to a black level.
15. The system of claim 11 , wherein the operations further comprise:
approximating the identified contours as polygons; and
wherein the operation of extracting the card image comprises:
extracting the card image as the region enclosed by the area as a minimum rectangular perimeter encompassing all the polygons.
16. The system of claim 15 , wherein the operations further comprise:
counting modified pixels for at least one boundary of the area; and
based on determining that a number of the modified pixels exceeds a predetermined threshold, moving the at least one boundary inward to generate an adjusted area.
17. The system of claim 16 , wherein the operations further comprise:
validating the adjusted area by:
counting a first number of pixels in the video frame region;
counting a second number of pixels in the area;
counting a third number of pixels in the adjusted area; and
determining that the adjusted area is valid based on comparing the first number of pixels, the second number of pixels, and the third number of pixels.
18. The system of claim 11 , wherein the operations further comprise:
generating a metadata associated with the video frame based analyzing the extracted card image.
19. The system of claim 18 , wherein the video frame is within a highlight of a video stream of a sporting event broadcast, and wherein the metadata is descriptive of a status of the sporting event during the highlight.
20. A non-transitory storage medium storing computer program instructions that when executed by a processor perform operations comprising:
selecting a video frame region that forms a portion of the video frame;
modifying pixel values of a set of pixels adjacent to a boundary of the video frame region;
removing background from the video frame region based on the modified pixel values;
generating an edge image based on the video frame region with removed background;
identifying contours in the edge image; and
extracting a card image from the video frame as a region enclosed by an area encompassing the contours.