Encoding a transparency (ALPHA) channel in a video bitstream
First Claim
Patent Images
1. A method to decode an image data stream, the method comprising:
- receiving, by a decoder, an image data stream having one or more syntax structures corresponding to an encoded foreground image and one or more syntax structures corresponding to an encoded transparency mask, wherein the one or more syntax structures corresponding to the encoded transparency mask are delimited from and are located in the image data stream after the one or more syntax structures corresponding to the foreground image in the image data stream;
decoding, by the decoder, the one or more syntax structures corresponding to the foreground image to produce a decoded foreground image;
decoding, by the decoder, the one or more syntax structures corresponding to the transparency mask to produce a decoded transparency mask; and
developing a display image as a function of the decoded foreground image and the decoded transparency mask.
0 Assignments
0 Petitions
Accused Products
Abstract
Disclosed herein is a technique for delimiting the alpha channel at the NAL layer in codecs like H.264 to facilitate the optional nature of the alpha channel. In coded video sequences that include alpha, there is one alpha picture for every primary coded (e.g., luma-chroma) picture, and the coded alpha picture is contained in the same access unit as its corresponding primary coded picture. The alpha coded slice NAL units of each access unit are sent after the NAL units of the primary coded picture and redundant coded pictures, if any. The presence or absence of the alpha NAL units does not affect the decoding of the remaining NAL units in any way.
-
Citations
23 Claims
-
1. A method to decode an image data stream, the method comprising:
-
receiving, by a decoder, an image data stream having one or more syntax structures corresponding to an encoded foreground image and one or more syntax structures corresponding to an encoded transparency mask, wherein the one or more syntax structures corresponding to the encoded transparency mask are delimited from and are located in the image data stream after the one or more syntax structures corresponding to the foreground image in the image data stream; decoding, by the decoder, the one or more syntax structures corresponding to the foreground image to produce a decoded foreground image; decoding, by the decoder, the one or more syntax structures corresponding to the transparency mask to produce a decoded transparency mask; and developing a display image as a function of the decoded foreground image and the decoded transparency mask. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
10. A method to decode an image data stream, the method comprising:
-
extracting, by a decoder, a syntax structure corresponding to an encoded foreground image from an image data stream; extracting, by the decoder, a syntax structure corresponding to an encoded transparency mask from the image data stream, wherein the syntax structure corresponding to the transparency mask is delimited from and is located in the image data stream after the syntax structure corresponding to the foreground image in the image data stream; decoding, by the decoder, the syntax structure corresponding to the encoded foreground image to generate a decoded foreground image; decoding, by the decoder, the syntax structure corresponding to the encoded transparency mask to generate a decoded transparency mask; and developing, by the decoder, a display image as a function of the decoded foreground image and the decoded transparency mask.
-
-
11. A method to encode images, the method comprising:
-
encoding, by an encoder, a foreground image into one or more syntax structures; encoding, by the encoder, a transparency mask into one or more syntax structures; combining, by the encoder, the one or more syntax structures corresponding to the encoded foreground image and the one or more syntax structures corresponding to the encoded transparency mask into an image data stream, wherein the one or more syntax structures corresponding to the encoded transparency mask are delimited from and are located in the image data stream after the one or more syntax structures corresponding to the encoded foreground image in the image data stream. - View Dependent Claims (12, 13, 14, 15, 16, 17)
-
-
18. A method to decode an image comprising:
-
receiving, by a decoder, an image data stream encoded in accordance with ITU Recommendation H.264, the image data stream including a primary encoded picture, a delimiter following the primary encoded picture in the image data stream and an encoded transparency picture following the delimiter in the image data stream; decoding, by the decoder, the encoded primary picture to produce a decoded primary picture; decoding, by the decoder, the encoded transparency picture to produce a decoded transparency picture; and developing, by the decoder, a display image as a function of the decoded primary picture and the decoded transparency picture. - View Dependent Claims (19, 20, 21, 22, 23)
-
Specification