GENERATING AN OUTPUT FRAME FOR INCLUSION IN A VIDEO SEQUENCE

摘要

In frame rate conversion, output frames (e.g. interpolated frames) are generated for inclusion in a video sequence. A first motion-compensated image may be generated for representing an output frame, e.g. by using motion estimation based on one or more of the existing frames of the video sequence. At least part of the first motion-compensated image is smoothed to determine a smoothed motion-compensated image. Trust indications can be determined for the pixels of the first motion-compensated image to indicate levels of trust in the pixel values. The trust indications may be used to determine how to selectively combine the pixels of the first motion-compensated image and the pixels of the smoothed motion-compensated image to thereby generate the output frame.

主权项

1. A method of generating an output frame for inclusion in a video sequence that comprises a plurality of existing frames, the method comprising:
using at least one prediction based on at least one of the existing frames of the video sequence to generate a first motion-compensated image for representing the output frame; determining trust indications for pixels of the first motion-compensated image, said trust indications indicating levels of trust in the values of said pixels of the first motion-compensated image; smoothing at least part of the determined first motion-compensated image to determine a smoothed motion-compensated image; and generating the output frame by using the determined trust indications for the pixels of the first motion-compensated image to selectively combine the first motion-compensated image and the smoothed motion-compensated image.