| CPC H04N 19/139 (2014.11) [H04N 19/119 (2014.11); H04N 19/14 (2014.11)] | 20 Claims |

|
1. A computer-implemented method for processing a video file, the method comprising:
identifying a plurality of still images from an original video file;
identifying, based on a geometric framework applicable to respective ones of the plurality of still images, two or more first areas within a first still image of the plurality of still images;
identifying, based on the geometric framework, two or more second areas within a second still image of the plurality of still images, wherein the two or more second areas correspond to the two or more first areas;
identifying, based on the geometric framework, two or more third areas within a third still image of the plurality of still images, wherein the two or more third areas correspond to the two or more second areas;
determining, for each one of the two or more first areas relative to each corresponding one of the two or more second areas, a first value associated with a difference between the first still image and the second still image;
determining, for each one of the two or more second areas relative to each corresponding one of the two or more third areas, a second value associated with a difference between the second still image and the third still image;
determining, based on the first values, a first preferred subframe for a first segment associated with the first still image and the second still image;
determining, based on the second values, a second preferred subframe for a second segment associated with the second still image and the third still image;
modifying, based on the first preferred subframe, the first still image through one or more of (i) a cropping process, (ii) a zooming process, or (iii) a panning process to create a modified first still image;
modifying, based on the second preferred subframe, the second still image through one or more of (i) a cropping process, (ii) a zooming process, or (iii) a panning process to create a modified second still image; and
compiling the modified first still image and the modified second still image into a new video file.
|