Apparatus and method for data partitioning to improving error resilience
AssigneesSarnoff Corporation, Sharp Kabushiki Kaisha
InventorsBing-Bing Chai, Iraj Sodagar
Life of this patent
Estimating the term…
Abstract
An apparatus and a method for partitioning data to improve error resilience. Specifically, one or more segment markers (symbols) are entropy encoded along with the bitstream (payload) into a packet.
Claims19
A method for packetizing a bitstream, where said bitstream carries an entropy encoded image, said method comprising the steps of:
- a) generating a packet header;
- b) generating a payload comprising a plurality of payload segments with each payload segment having at least one sub-unit comprising of hierarchical subband decomposed coefficients of said entropy encoded image;
- and
- c) inserting a coded segment marker after each of said plurality of payload segments.
A data structure stored on a computer readable medium comprising:
- a packet header;
- a payload, coupled to said packet header, where said payload comprises a plurality of payload segments, with each of said payload segments having at least one sub-unit comprising of hierarchical subband decomposed coefficients of an entropy encoded image;
- and
- a plurality of segment markers with one of said plurality of segment markers being coupled after one of said payload segments.
A method for decoding a bitstream, where said bitstream carries an entropy encoded image, said method comprising the steps of:
- a) decoding a packet header;
- b) decoding a payload comprising a plurality of payload segments with each payload segment having at least one sub-unit of said entropy encoded image;
- c) searching for a coded segment marker after each of said plurality of payload segments;
- and
- d) deleting said at least one sub-unit of said entropy encoded image in one of said plurality of payload segments if said coded segment marker after a corresponding payload segment is undetected.
A computer-readable medium having stored thereon a plurality of instructions, the plurality of instructions including instructions which, when executed by a processor, cause the processor to perform the steps comprising of:
- a) generating a packet header;
- b) generating a payload comprising a plurality of payload segments with each payload segment having at least one sub-unit comprising of hierarchical subband decomposed coefficients of said entropy encoded image;
- and
- c) inserting a coded segment marker after each of said plurality of payload segments.
A computer-readable medium having stored thereon a plurality of instructions, the plurality of instructions including instructions which, when executed by a processor, cause the processor to perform a method for decoding a bitstream, where said bitstream carries an entropy encoded image, the steps comprising of:
- a) decoding a packet header;
- b) decoding a payload comprising a plurality of payload segments with each payload segment having at least one sub-unit of said entropy encoded image;
- c) searching for a coded segment marker after each of said plurality of payload segments;
- and
- d) deleting said at least one sub-unit of said entropy encoded image in one of said plurality of payload segments if said coded segment marker after a corresponding payload segment is undetected.
A method for entropy coding, said method comprising the steps of:
- (a) obtaining a plurality of coefficients representative of an image for entropy coding, wherein each of said plurality of coefficients has relative coordinate (i, j);
- and
- (b) entropy coding a current coefficient in accordance with a context model consisting only three neighboring coefficients at relative coordinates of (i−1, j−1), (i−1, j) and (i, j−1), wherein said entropy coding step (b) codes said coefficients using a context model in accordance with:
- Model_no=f(i−1, j−1)+f(i−1, j)*2+f(i, j−1)*4
- where f ( x , y ) = { 1 , if coeff ( x , y ) is available and nonzero 0 , else
- Model_no is the number of a context model selected, coeff(x, y) is a value of a coefficient at coordinate (x, y), and f(i, j) is a function f(x, y) at coordinate (i, j).
Dependencies are inferred from the claim text.