IPC분류정보
국가/구분 |
United States(US) Patent
등록
|
국제특허분류(IPC7판) |
|
출원번호 |
US-0398741
(2012-02-16)
|
등록번호 |
US-8494295
(2013-07-23)
|
발명자
/ 주소 |
- Chen, Wen-hsiung
- Wu, Fang
- Toebes, John A.
- Rosenfeld, Dov
- Pelton, Gregory D.
|
출원인 / 주소 |
|
대리인 / 주소 |
|
인용정보 |
피인용 횟수 :
0 인용 특허 :
94 |
초록
▼
Disclosed are apparatuses, computer readable storage media with instructions that carry out methods, and methods of (a) coding quantized transform coefficients as occur in image compression to form codewords, and (b) decoding a bitstream comprising the codewords. The coding combines a position ident
Disclosed are apparatuses, computer readable storage media with instructions that carry out methods, and methods of (a) coding quantized transform coefficients as occur in image compression to form codewords, and (b) decoding a bitstream comprising the codewords. The coding combines a position identifying method to identify the relative position of clusters of consecutive non-zero-valued quantized transform coefficients, an amplitude event identifying method to identify amplitude events in the clusters of non-zero-valued quantized transform coefficients, and a coding method to code the position events, amplitude events, and signs of the amplitudes in the clusters. The coding method is particularly applicable to an ordered series of quantized transform coefficients where clusters are likely-to-occur.
대표청구항
▼
1. A method of operating processing hardware, the method comprising: (A) recognizing in a bitstream of codewords one or more codewords for a particular ordered series of quantized transform coefficients of a particular block of image data, the bitstream representing a plurality of ordered series of
1. A method of operating processing hardware, the method comprising: (A) recognizing in a bitstream of codewords one or more codewords for a particular ordered series of quantized transform coefficients of a particular block of image data, the bitstream representing a plurality of ordered series of quantized transform coefficients of blocks of image data, each ordered series corresponding to one block of the image data, each quantized transform coefficient having a respective amplitude from a finite set of amplitudes consisting of a most likely-to-occur amplitude and at least one other amplitude, the at least one other amplitude including a second-most likely-to-occur amplitude, the one or more codewords identifying the position in the particular ordered series of any cluster of at least one quantized coefficient having other than the most likely-to-occur amplitude, the one or more codewords further defining the amplitudes of the quantized coefficients having amplitudes other than the most likely-to-occur amplitude,wherein the codewords for the particular ordered series are determined from the particular ordered series of quantized transform coefficients by a position event identifying method to identify for each series representing a block of image data, one or more position events that each includes a cluster of one or more consecutive quantized coefficients having other than the most likely-to-occur amplitude preceded by either no quantized coefficients having the most likely-to-occur amplitude, or a run of one or more consecutive quantized coefficients having the most likely-to-occur amplitude, each position event identified by the runlengths of the preceding run of consecutive quantized coefficients having the most likely-to-occur amplitude, including a runlength of 0, and the runlength of the cluster of consecutive quantized coefficients having other than the most likely-to-occur amplitude, andan amplitude-event-identifying method to identify amplitude events defining for each series representing a block of image data the amplitudes of the quantized coefficients having amplitudes other than the most likely-to-occur amplitude, anda coding method that forms codewords for the position events and amplitude events;wherein the coding of the coding method is characterized in that for at least some identified position and amplitude events, the more likely to occur the position and amplitude events, the shorter the codewords that encode the position and amplitude events; and(B) determining from the recognized codewords the position in the particular ordered series of any cluster of quantized coefficients having other than the most likely-to-occur amplitude and the amplitude of each quantized coefficients in the particular series. 2. A method as recited in claim 1, wherein the quantized transform coefficients are formed using a transform configured such that the most likely-to-occur amplitude is 0, and the second-most likely-to-occur amplitude is 1, such that the position event identifying method identifies events that include a cluster of one or more non-zero-valued coefficients, and wherein the quantized transform coefficients of each block of image data includes a DC term, and wherein the DC term is separately encoded, such that the particular ordered series consists of the non-DC quantized transform coefficients of the particular block of image data. 3. A method as recited in claim 1, wherein the quantized transform coefficients are formed using a transform configured such that the most likely-to-occur amplitude is 0, and the second-most likely-to-occur amplitude is 1, such that the position event identifying method identifies events that include a cluster of one or more non-zero-valued coefficients, and the coding method includes separately encoding the signs of the amplitudes. 4. A method as recited in claim 1, wherein the quantized transform coefficients are formed using a transform configured such that the most likely-to-occur amplitude is 0, and the second-most likely-to-occur amplitude is 1, such that the position event identifying method identifies events that include a cluster of one or more non-zero-valued coefficients, and wherein the coding method that encodes the amplitude events and the position events uses coding tables. 5. A method as recited in claim 4, wherein each of the coding tables is constructed by assigning a variable length code to at least some of the events being coded by the coding table according to measured or assumed statistics of the respective event being coded by the coding table. 6. A method as recited in claim 1, wherein the amplitude-event-identifying method includes identifying amplitude events that include a run of consecutive quantized coefficients having amplitude other than the most likely-to-occur amplitude. 7. A method as recited in claim 1, wherein the amplitude-event-identifying method uses a method selected from a set consisting of a first amplitude-event-identifying method and a second amplitude-event-identifying method, wherein the first amplitude-event-identifying method comprises: in each cluster of consecutive quantized coefficients having other than the most likely-to-occur amplitude, identifying amplitude events of at least one consecutive quantized coefficient, an identified amplitude event including: a sequence of at least one consecutive quantized coefficient having the second-most likely-to-occur amplitude that ends in a quantized coefficient having other than the second-most likely-to-occur amplitude,an assertion that the amplitude event has only a quantized coefficient having other than the second-most likely-to-occur amplitude, oran assertion that the amplitude event has only a sequence of consecutive quantized coefficients having the second-most likely-to-occur amplitude,wherein each identified amplitude event is defined by the length of the cluster of consecutive quantized coefficients having amplitude other than the most likely-to-occur amplitude, the length of the sequence of at least one consecutive quantized coefficient having the second-most likely-to-occur amplitude, and by the amplitude of the last quantized coefficient in the run, andthe second amplitude-event-identifying method including identifying, in each run of consecutive quantized coefficients of value or values having other than the most likely-to-occur amplitude, amplitude events of a single quantized coefficient having other than the most likely-to-occur amplitude. 8. A method as recited in claim 1, wherein the amplitude-event-identifying method identifies individual amplitudes in the clusters. 9. A method as recited in claim 1, wherein the coding method includes jointly encoding identified position events with identified amplitude events. 10. A method as recited in claim 1, wherein the coding method separately encodes the identified position events and the identified amplitude events. 11. A method as recited in claim 1, wherein the identifying amplitude events includes identifying amplitude events that include a run of consecutive quantized coefficients having amplitude other than the most likely-to-occur amplitude. 12. A method as recited in claim 1, wherein the amplitude event identifying is selected from the set consisting of a first method of identifying and a second method of identifying, wherein the first method of identifying comprises: in any cluster of consecutive quantized coefficients having other than the most likely-to-occur amplitude, identifying amplitude events of at least one consecutive quantized coefficient, an identified amplitude event including: a sequence of at least one consecutive quantized coefficient having the second-most likely-to-occur amplitude that ends in a quantized coefficient having other than the second-most likely-to-occur amplitude,an assertion that the amplitude event has only a quantized coefficient having other than the second-most likely-to-occur amplitude, oran assertion that the amplitude event has only a sequence of consecutive quantized coefficients having the second-most likely-to-occur amplitude,wherein each identified amplitude event defined by the length of the cluster of consecutive quantized coefficients having amplitude other than the most likely-to-occur amplitude, the length of the sequence of at least one consecutive quantized coefficient having the second-most likely-to-occur amplitude, and by the amplitude of the last quantized coefficient in the run, andthe second method of identifying comprising identifying, in each run of consecutive quantized coefficients of value or values having other than the most likely-to-occur amplitude, amplitude events of a single quantized coefficient having other than the most likely-to-occur amplitude. 13. A method as recited in claim 1, wherein the amplitude event identifying identifies individual amplitudes in the clusters. 14. A computer-readable hardware storage medium configured with instructions that when executed by at least one processor of a processing system cause the processing system to carry out a method of decoding a bitstream of codewords, the method comprising: (A) recognizing in the bitstream one or more codewords for a particular ordered series of quantized transform coefficients of a particular block of image data, the bitstream representing a plurality of ordered series of quantized transform coefficients of blocks of image data, each ordered series corresponding to one block of the image data, each quantized transform coefficient having a respective amplitude from a finite set of amplitudes consisting of a most likely-to-occur amplitude and at least one other amplitude, the at least one other amplitude including a second-most likely-to-occur amplitude, the one or more codewords identifying the position in the particular ordered series of any cluster of at least one quantized coefficient having other than the most likely-to-occur amplitude, the one or more codewords further defining the amplitudes of the quantized coefficients having amplitudes other than the most likely-to-occur amplitude,wherein the codewords are determined from the particular ordered series of quantized transform coefficients by a position event identifying method to identify for each series representing a block of image data, one or more position events that each includes a cluster of one or more consecutive quantized coefficients having other than the most likely-to-occur amplitude preceded by either no quantized coefficients having the most likely-to-occur amplitude, or a run of one or more consecutive quantized coefficients having the most likely-to-occur amplitude, each position event identified by the runlengths of the preceding run of consecutive quantized coefficients having the most likely-to-occur amplitude, including a runlength of 0, and the runlength of the cluster of consecutive quantized coefficients having other than the most likely-to-occur amplitude, andan amplitude-event-identifying method to identify amplitude events defining for each series representing a block of image data the amplitudes of the quantized coefficients having amplitudes other than the most likely-to-occur amplitude, anda coding method that encodes the position events and amplitude events to form codewords, wherein the coding method is characterized in that for at least some identified position and amplitude events, the more likely to occur the position and amplitude events, the shorter the position and amplitude codewords that encode the position and amplitude events;(B) determining from the recognized codewords the position in the particular ordered series of any cluster of quantized coefficients having other than the most likely-to-occur amplitude and the amplitude of each quantized coefficients in the particular series. 15. A computer-readable hardware storage medium as recited in claim 14, wherein the quantized transform coefficients are formed using a transform configured such that the most likely-to-occur amplitude is 0, and the second-most likely-to-occur amplitude is 1, such that the position event identifying method identifies events that include a cluster of one or more non-zero-valued coefficients, and wherein the coding method includes jointly encoding identified position events with identified amplitude events. 16. A computer-readable hardware storage medium as recited in claim 15, wherein the amplitude-event-identifying method uses a method selected from a set consisting of a first amplitude-event-identifying method and a second amplitude-event-identifying method, wherein the first amplitude-event-identifying method comprises: in each cluster of consecutive quantized coefficients having other than the most likely-to-occur amplitude, identifying amplitude events of at least one consecutive quantized coefficient, an identified amplitude event including:a sequence of at least one consecutive quantized coefficient having the second-most likely-to-occur amplitude that ends in a quantized coefficient having other than the second-most likely-to-occur amplitude,an assertion that the amplitude event has only a quantized coefficient having other than the second-most likely-to-occur amplitude, oran assertion that the amplitude event has only a sequence of consecutive quantized coefficients having the second-most likely-to-occur amplitude,wherein each identified amplitude event is defined by the length of the cluster of consecutive quantized coefficients having amplitude other than the most likely-to-occur amplitude, the length of the sequence of at least one consecutive quantized coefficient having the second-most likely-to-occur amplitude, and by the amplitude of the last quantized coefficient in the run, andthe second amplitude-event-identifying method comprising identifying, in each run of consecutive quantized coefficients of value or values having other than the most likely-to-occur amplitude, amplitude events of a single quantized coefficient having other than the most likely-to-occur amplitude. 17. A computer-readable hardware storage medium as recited in claim 14, wherein the quantized transform coefficients are formed using a transform configured such that the most likely-to-occur amplitude is 0, and the second-most likely-to-occur amplitude is 1, such that the position event identifying method identifies events that include a cluster of one or more non-zero-valued coefficients, and wherein the coding method includes separately encoding the identified position events and the identified amplitude events. 18. A computer-readable hardware storage medium as recited in claim 14, wherein the amplitude-event-identifying method uses Multi-Dimensional Amplitude Event Identifying, Multi-Dimensional Amplitude Event Identifying including identifying amplitude events that include a run of consecutive signals of amplitude or amplitudes other than the most likely-to-occur amplitude. 19. An apparatus operative to decode a bitstream of codewords, the apparatus comprising: one or more processors; anda storage subsystem configured with instructions that when executed by at least one of the processors cause the apparatus to: (A) recognize in the bitstream one or more codewords for a particular ordered series of quantized transform coefficients of a particular block of image data, the bitstream representing a plurality of ordered series of quantized transform coefficients of blocks of image data, each ordered series corresponding to one block of the image data, each quantized transform coefficient having a respective amplitude from a finite set of amplitudes consisting of a most likely-to-occur amplitude and at least one other amplitude, the at least one other amplitude including a second-most likely-to-occur amplitude, the one or more codewords identifying the position in the particular ordered series of any cluster of at least one quantized coefficient having other than the most likely-to-occur amplitude, the one or more codewords further defining the amplitudes of the quantized coefficients having amplitudes other than the most likely-to-occur amplitude,wherein the codewords are determined from the particular ordered series of quantized transform coefficients by a position event identifying method to identify for each series representing a block of image data, one or more position events that each includes a cluster of one or more consecutive quantized coefficients having other than the most likely-to-occur amplitude preceded by either no quantized coefficients having the most likely-to-occur amplitude, or a run of one or more consecutive quantized coefficients having the most likely-to-occur amplitude, each position event identified by the runlengths of the preceding run of consecutive quantized coefficients having the most likely-to-occur amplitude, including a runlength of 0, and the runlength of the cluster of consecutive quantized coefficients having other than the most likely-to-occur amplitude, andan amplitude-event-identifying method to identify amplitude events defining for each series representing a block of image data the amplitudes of the quantized coefficients having amplitudes other than the most likely-to-occur amplitude, anda coding method that encodes the position events and amplitude events to form codewords, wherein the coding method is characterized in that for at least some identified position and amplitude events, the more likely to occur the position and amplitude events, the shorter the position and amplitude codewords that encode the position and amplitude events;(B) determine from the recognized codewords the position in the particular ordered series of any cluster of quantized coefficients having other than the most likely-to-occur amplitude and the amplitude of each quantized coefficients in the particular series. 20. An apparatus as recited in claim 19, wherein the quantized transform coefficients are formed using a transform configured such that the most likely-to-occur amplitude is 0, and the second-most likely-to-occur amplitude is 1, such that the position event identifying method identifies events that include a cluster of one or more non-zero-valued coefficients, and wherein the coding method includes jointly encoding identified position events with identified amplitude events. 21. An apparatus as recited in claim 19, wherein the quantized transform coefficients are formed using a transform configured such that the most likely-to-occur amplitude is 0, and the second-most likely-to-occur amplitude is 1, such that the position event identifying method identifies events that include a cluster of one or more non-zero-valued coefficients, and wherein the coding method includes separately encoding the identified position events and the identified amplitude events. 22. An apparatus as recited in claim 19, wherein the amplitude-event-identifying method uses Multi-Dimensional Amplitude Event Identifying, Multi-Dimensional Amplitude Event Identifying including identifying amplitude events that include a run of consecutive signals of amplitude or amplitudes other than the most likely-to-occur amplitude. 23. An apparatus as recited in claim 19, wherein the amplitude-event-identifying method uses a method selected from a set consisting of a first amplitude-event-identifying method and a second amplitude-event-identifying method, wherein the first amplitude-event-identifying method comprises: in each cluster of consecutive quantized coefficients having other than the most likely-to-occur amplitude, identifying amplitude events of at least one consecutive quantized coefficient, an identified amplitude event including: a sequence of at least one consecutive quantized coefficient having the second-most likely-to-occur amplitude that ends in a quantized coefficient having other than the second-most likely-to-occur amplitude,an assertion that the amplitude event has only a quantized coefficient having other than the second-most likely-to-occur amplitude, oran assertion that the amplitude event has only a sequence of consecutive quantized coefficients having the second-most likely-to-occur amplitude,wherein each identified amplitude event is defined by the length of the cluster of consecutive quantized coefficients having amplitude other than the most likely-to-occur amplitude, the length of the sequence of at least one consecutive quantized coefficient having the second-most likely-to-occur amplitude, and by the amplitude of the last quantized coefficient in the run, andthe second amplitude-event-identifying method comprising identifying, in each run of consecutive quantized coefficients of value or values having other than the most likely-to-occur amplitude, amplitude events of a single quantized coefficient having other than the most likely-to-occur amplitude. 24. A method of operating processing hardware to encode a plurality of ordered series of quantized transform coefficients, the method comprising: for each ordered series in the plurality, each ordered series being of quantized transform coefficients of a block of image data, each quantized transform coefficient having a respective amplitude from a finite set of amplitudes consisting of a most likely-to-occur amplitude and at least one other amplitude, the at least one other amplitude including a second-most likely-to-occur amplitude:accepting the ordered series of quantized transform coefficients of the block of image data; identifying in the ordered series any position event, each position event defining the position in the ordered series of a cluster of one or more consecutive quantized coefficients having other than the most likely-to-occur amplitude preceded by either no quantized coefficients having the most likely-to-occur amplitude, or a run of one or more consecutive quantized coefficients having the most likely-to-occur amplitude, each position event identified by the runlengths of the preceding run of consecutive quantized coefficients having the most likely-to-occur amplitude, including a runlength of 0, and the runlength of the cluster of consecutive quantized coefficients having other than the most likely-to-occur amplitude;identifying amplitude events defining the amplitudes of the quantized coefficients in the particular ordered series having amplitudes other than the most likely-to-occur amplitude;encoding the identified position events and the identified amplitude events to form codewords, wherein the encoding of the identified position and the amplitude events is characterized in that, for at least some identified position and amplitude events in the plurality of ordered series, the more likely to occur the position and amplitude events, the shorter the codewords that encode the identified position and amplitude events.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.