Encoding device and method

1. An encoding device for outputting encoded data, based on a weblet co-efficient obtained by weblet transform, comprising:

a type determination unit for inputting each bit of the weblet co-efficient in a depth direction in parallel and determining type information indicating which the bit is, the first “1” bit when viewed from MSB side, a bit located further on the MSB side than the first “1” bit or a bit further located on the LSB side than the first “1” bit;

a buffer unit for buffering the determined type information of each bit for each bit depth;

a pass determination unit for determining pass information of a type information to process, of a plurality of pieces of type information, based on the plurality of pieces of buffered type information of the same depth;

a context generation unit for generating a context, based on the type information outputted from the buffer unit and the pass information determined for the outputted type information; and

an arithmetic encoding unit for performing arithmetic encoding, based on the generated context.

2. The encoding device according to claim 1 , wherein

the pass determination unit generates significant states of eight neighbor bits of the type information to process, based on type information of the eight neighbor bits of type information to process and pass information of bits processed before the type information to process, of its eight neighbor bits, and

determines pass information of the type information to process, based on the generated eight neighbor significant states and the type information to process.

3. The encoding device according to claim 1 , wherein

the pass determination unit determines pass information of the type information to process to be a magnitude refinement (MR) pass if the type information to process is the type information indicating the bit (corresponding to the type information to process) is located further on a LSB side than the first “1” bit when viewed from an MSB side,

determines pass information of the type information to process to be a clean-up (CL) pass if the type information to process is the first “1” bit when viewed from an MSB side and the eight neighbor significant states of the type information to process all are non-significant, and

determines pass information of the type information to process to be a significance propagation (SP) pass if the type information to process is the first “1” bit when viewed from an MSB side and at least one of the eight neighbor significant states of the type information to process is significant.

4. The encoding device according to claim 1 , wherein

the type determination unit is provided for each corresponding stripe of each bit plane constituting a code block,

the buffer unit, the pass determination unit and the context generation unit are provided for each bit plane and also for each stripe in the bit plane, and

a starting time of a process of inputting each bit of a weblet co-efficient in a depth direction in parallel, according to a position in a bit plane of the stripe is slid.

5. The encoding device according to claim 4 , wherein

If pre-process bits of the bit corresponding to the type information processed in the pass determination unit, are obtained from a stripe immediately before a stripe corresponding to the pass determination unit when calculating eight neighbor significant states of the type information, the staring time of the process of the type determination unit is slid for each stripe in such a way that pass information of the pre-process bits are already determined in a stripe immediately before the relevant stripe.

6. The encoding device according to claim 1 , wherein

the type determination unit inputs each bit of the weblet co-efficient in a depth direction in parallel and determines type information indicating which the bit is, the first “1” bit when viewed from an MSB side, a bit located further on the MSB side than the first “1” bit, a bit located further on an LSB side than the first “1” bit by one bit or a bit located further on the LSB side than the LSB side bit by one bit, and

the context generation unit generates a context of the first MR pass, using type information indicating that the bit is the bit located further on an LSB side than the first “1” bit by one bit.

7. An encoding device for outputting encoded data, based on a weblet co-efficient indicated by a significant state variable obtained by weblet transform, comprising:

a type determination unit for inputting each bit of the weblet co-efficient in a depth direction in parallel and determining type information indicating which the bit is, the first significant bit when viewed from MSB side, a bit located further on the MSB side than the first significant bit or a bit further located on the LSB side than the first significant bit;

a buffer unit for buffering the determined type information of each bit for each bit depth;

a pass determination unit for determining pass information of a type information to process, of a plurality of pieces of type information, based on the plurality of pieces of buffered type information of the same depth;

a context generation unit for generating a context, based on the type information outputted from the buffer unit and the pass information determined for the outputted type information; and

an arithmetic encoding unit for performing arithmetic encoding, based on the generated context.

8. An encoding method for outputting encoded data, based on a weblet co-efficient obtained by weblet transform, comprising:

a type determination step of inputting each bit of the weblet co-efficient in a depth direction in parallel and determining type information indicating which the bit is, the first “1” bit when viewed from MSB side, a bit located further on the MSB side than the first “1” bit or a bit further located on the LSB side than the first “1” bit;

a buffering step of buffering the determined type information of each bit for each bit depth;

a pass determination step of determining pass information of a type information to process, of a plurality of pieces of type information, based on the plurality of pieces of buffered type information of the same depth;

a context generation step of generating a context, based on the type information outputted from the buffer unit and the pass information determined for the outputted type information; and

an arithmetic encoding step of performing arithmetic encoding, based on the generated context.

9. The encoding method according to claim 8 , wherein

in the pass determination step, significant states of eight neighbor bits of the type information to process are generated, based on type information of the eight neighbor bits of type information to process and pass information of bits processed before the type information to process, of its eight neighbor bits, and

pass information of the type information to process is determined, based on the generated eight neighbor significant states and the type information to process.

10. The encoding method according to claim 8 , wherein

in the pass determination step, if the type information to process is the type information indicating the bit (corresponding to the type information to process) is located further on a LSB side than the first “1” bit when viewed from an MSB side, pass information of the type information to process to be a MR pass,

if the type information to process is the first “1” bit when viewed from an MSB side and the eight neighbor significant states of the type information to process all are non-significant, pass information of the type information to process is determined to be a CL pass, and

if the type information to process is the first “1” bit when viewed from an MSB side and at least one of the eight neighbor significant states of the type information to process is significant, pass information of the type information to process is determined to be a SP pass.

11. The encoding method according to claim 8 , wherein

the type determination step is executed for each corresponding stripe of each bit plane constituting a code block,

the buffering step, the pass determination step and the context generation step are executed for each bit plane and also for each stripe in the bit plane, and

in the type determination step, a starting time of a process of inputting each bit of a weblet co-efficient in a depth direction in parallel is slid according to a position in a bit plane of the stripe.

12. The encoding method according to claim 11 , wherein

If pre-process bits of the bit corresponding to the type information processed in the pass determination step, are obtained from a stripe immediately before a stripe corresponding to the pass determination step when calculating eight neighbor significant states of the type information, the staring time of the process of the type determination step is slid for each stripe in such a way that pass information of the pre-processed bits are already determined in a stripe immediately before the relevant stripe.

13. The encoding method according to claim 8 , wherein

in the type determination step, each bit of the weblet co-efficient is inputted in a depth direction in parallel and type information indicating which the bit is, the first “1” bit when viewed from an MSB side, a bit located further on the MSB side than the first “1” bit, a bit located further on an LSB side than the first “1” bit by one bit or a bit located further on the LSB side than the LSB side bit by one bit is determined, and

in the context generation step, a context of the first MR pass is generated, using type information indicating that the bit is the bit located further on an LSB side than the first “1” bit by one bit.