System and method for video encoding using adaptive segmentation

A method for creating a stream of encoded video information including a plurality of frames each having a plurality of blocks, comprising:

• for at least a first frame, creating a segment map associating at least one block in the first frame with at least one segment identifier, and associating the at least one segment identifier with at least one segment parameter;
• encoding the segment identifier using a binary arithmetic encoder;
• and
• encoding the at least one block using the at least one segment parameter.

A method for decoding a stream of encoded video information including a plurality of frames each having a plurality of blocks, comprising:

• for at least a first frame, reading from the video information a segment map associating at least one block in the first frame with at least one segment identifier, the segment identifier being decoded using a binary arithmetic decoder;
• reading from the encoded video information at least one segment parameter associated with the at least one segment identifier;
• decoding the at least one block using the at least one segment parameter;
• and
• for at least a second frame, decoding a block in the second frame corresponding to the at least one block in the first frame using the at least one segment identifier.