Embedding frame masks in a video stream

A computer-implemented method comprising:

• at a video encoder:
• encoding a frame of video data comprising an array of pixels to generate an encoded video frame, wherein the array of pixels comprise foreground pixels and background pixels;
• and
• transmitting the encoded video frame;
• wherein the foreground pixels have respective original luma values,
• wherein the original luma values are bounded within a first luma range;
• wherein encoding the frame of video data comprises converting the original luma values of the foreground pixels to updated luma values by applying a mapping function, wherein the updated luma values are bounded within a second luma range, and wherein the second luma range is shifted and/or
• compressed from the first luma range using the mapping function;
• at a video decoder:
• decoding the encoded video frame to obtain decoded luma values corresponding to the array of pixels;
• extracting a foreground mask based on the decoded luma values, wherein the foreground mask identifies a subset of the array of pixels to be in a foreground of the frame;
• and
• converting the decoded luma values of the subset of the array of pixels identified by the foreground mask to new luma values by applying an inverse mapping function to the decoded luma values of the subset of the array of pixels identified by the foreground mask, wherein the inverse mapping function reverses the mapping function, and wherein the new luma values are bounded within the first luma range.

A computer-implemented method comprising:

• receiving a frame of video data comprising an array of pixels having respective encoded luma values, wherein the encoded luma values are converted from original luma values by applying a mapping function, wherein the original luma values are bounded within a first luma range, wherein the encoded luma values are bounded within a second luma range, and wherein the second luma range is shifted and/or compressed from the first luma range using the mapping function;
• decoding the encoded luma values to obtain decoded luma values corresponding to the array of pixels;
• extracting a foreground mask based on the decoded luma values, wherein the foreground mask identifies a subset of the array of pixels to be in a foreground of the frame;
• and
• converting the decoded luma values of the subset of the array of pixels identified by the foreground mask to new luma values by applying an inverse mapping function to the decoded luma values of the subset of the array of pixels identified by the foreground mask, wherein the inverse mapping function reverses the mapping function, and wherein the new luma values are bounded within the first luma range.