Method and apparatus to quantize and dequantize input signal, and method and apparatus to encode and decode input signal

Abstract

A method and apparatus to quantize and dequantize an input signal, and a method and apparatus to encode and decode an input signal. The method of quantizing an input signal includes determining a quantization scale type according to a distribution feature of the input signal, and quantizing the input signal according to the determined quantization scale type. Accordingly, when a number of assigned bits is small in an encoding process, signal distortion can be minimized without an increase in complexity or having to use large sized additional information in order to determine an optimum scale. In addition, the input signal can be encoded by considering a trade-off of a distortion rate corresponding to the number of assigned bits of the input signal.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

These and/or other aspects and advantages of the present general inventive concept will become apparent and more readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:

FIG. 1 is a flowchart illustrating a method of quantizing an input signal according to an embodiment of the present general inventive concept;

FIG. 2 is a flowchart illustrating an operation of determining a quantization scale type in the method of FIG. 1;

FIG. 3 is a graph illustrating an example of an actual distribution of an input signal and a generalized Gaussian distribution similar to an actual distribution;

FIGS. 4A to 4D are graphs illustrating correlation between a number of assigned bits and a distortion rate according to generalized Gaussian distribution variables;

FIG. 5 illustrates an example of quantization scale types;

FIGS. 6A to 6D are graphs illustrating comparisons of a correlation between a number of assigned bits and a distortion rate according to a quantization method of the present general inventive concept as compared to a conventional quantization method;

FIG. 7 is a flowchart illustrating a method of encoding an input signal according to an embodiment of the present general inventive concept;

FIG. 8 is a flowchart illustrating an operation of determining a quantization scale type in the method of FIG. 7;

FIG. 9 is a flowchart illustrating a method of dequantizing an input signal according to an embodiment of the present general inventive concept;

FIG. 10 is a flowchart of a method of decoding an input signal according to an embodiment of the present general inventive concept;

FIG. 11 is a block diagram illustrating an apparatus to quantize an input signal according to an embodiment of the present general inventive concept;

FIG. 12 is a block diagram illustrating a type determination unit of the apparatus of FIG. 11;

FIG. 13 is a block diagram illustrating an apparatus to encode an input signal according to an embodiment of the present general inventive concept;

FIG. 14 is a block diagram illustrating a type determination unit of the apparatus of FIG. 13;

FIG. 15 is a block diagram illustrating an apparatus to dequantize an input signal according to an embodiment of the present general inventive concept; and

FIG. 16 is a block diagram illustrating an apparatus to decode an input signal according to an embodiment of the present general inventive concept.

Claims

1. A method of quantizing an input signal, the method comprising: determining a quantization scale type indicating a scale type for quantization according to a distribution feature of an input signal; andquantizing the input signal according to the determined quantization scale type and the distribution feature.

2. The method of claim 1, wherein the determining of the quantization scale type comprises: determining a generalized Gaussian distribution variable at which a generalized Gaussian distribution matches a distribution of the input signal;determining a correlation between the number of assigned bits used to encode the input signal and a distortion rate of the input signal by using the determined generalized Gaussian distribution variable; anddetermining the quantization scale type corresponding to the distortion rate of the input signal and determining a scale factor that is used for quantization from the determined correlation.

3. The method of claim 2, wherein the determining of the generalized Gaussian distribution variable comprises determining a minimum value of chi-square scale as the generalized Gaussian distribution variable.

4. The method of claim 2, wherein the determining of correlation comprises pre-storing correlation information respectively corresponding to the generalized Gaussian distribution variable.

5. The method of claim 2, wherein the determining of the quantization scale type comprises determining the quantization scale type having a minimum number of assigned bits with respect to the distortion rate of the input signal.

6. The method of claim 2, wherein the quantization scale type is determined from quantization scales having different quantization intervals with respect to a reference point of 0.

7. The method of claim 6, wherein the quantization scale type is defined by using Formula 1:

8. The method of claim 2, wherein the distortion rate of the input signal is determined by using psychoacoustic modeling.

9. The method of claim 2, wherein the determining of a quantization scale comprises determining the scale factor by using a standard deviation of the input signal.

10. The method of claim 2, wherein the quantizing of the input signal comprises calculating a normalization coefficient according to the determined quantization scale type, and quantizing the input signal by using the calculated normalization coefficient and the scale factor.

11. The method of claim 1, wherein the input signal is an audio signal.

12. A computer-readable medium having embodied thereon a computer program to execute a method of quantizing an input signal, the method comprising: determining a quantization scale type indicating a scale type for quantization according to a distribution feature of an input signal; andquantizing the input signal according to the determined quantization scale type and the distribution feature.

13. A method of encoding an input signal, comprising: determining a quantization scale type indicating a scale type for quantization according to a distribution feature of an input signal;quantizing the input signal according to the determined quantization scale type and the distribution feature; andencoding the quantized input signal.

14. The method of claim 13, wherein the determining of the quantization scale type comprises: determining a generalized Gaussian distribution variable at which a generalized Gaussian distribution matches a distribution of the input signal;determining correlation between a number of assigned bits used to encode the input signal and a distortion rate of the input signal by using the determined generalized Gaussian distribution variable; anddetermining a quantization scale type corresponding to the distortion rate of the input signal and determining a scale factor used for quantization from the determined correlation.

15. A method of dequantizing an input signal, comprising: dequantizing an input signal according to a quantization scale type corresponding to a distortion rate of the input signal and according to a scale factor used for quantization.

16. The method of claim 15, wherein the quantization scale type is determined by using a correlation between a number of assigned bits for encoding the input signal and the distortion rate of the input signal.

17. The method of claim 15, wherein the quantization scale type has a minimum number of bits with respect to the distortion rate of the input signal.

18. The method of claim 15, wherein the quantization scale type is determined from quantization scales having different quantization intervals with respect to a reference point of 0.

19. The method of claim 18, wherein the quantization scale type is defined by using Formula 1:

20. The method of claim 15, wherein the scale factor is determined by using a standard deviation of the input signal.

21. The method of claim 15, wherein the input signal is an audio signal.

22. A computer-readable medium having embodied thereon a computer program to execute a method of dequantizing an input signal, comprising: dequantizing the input signal according to a quantization scale type corresponding to a distortion rate of the input signal and according to a scale factor used for quantization.

23. A method of decoding an input signal, comprising: decoding an input signal; anddequantizing the decoded input signal according to a quantization scale type corresponding to a distortion rate of the decoded input signal and according to a scale factor used for quantization.

24. The method of claim 23, wherein the quantization scale type is determined from quantization scales having different quantization intervals with respect to a reference point of 0.

25. An apparatus to quantize an input signal, comprising: a type determination unit to determine a quantization scale type indicating a scale type for quantization according to a distribution feature of an input signal; anda quantizer to quantize the input signal according to the determined quantization scale type and the distribution feature.

26. The apparatus of claim 25, wherein the type determination unit comprises: a variable determination unit to determine a generalized Gaussian distribution variable at which a generalized Gaussian distribution matches a distribution of the input signal;a correlation determination unit to determine a correlation between a number of assigned bits used to encode the input signal and a distortion rate of the input signal by using the determined generalized Gaussian distribution variable; anda scale type and factor determination unit to determine a quantization scale type corresponding to the distortion rate of the input signal and to determine a scale factor used for quantization from the determined correlation.

27. The apparatus of claim 26, wherein the type determination unit further comprises a database to store correlation information corresponding to the generalized Gaussian distribution variable.

28. The apparatus of claim 26, wherein the scale type and factor determination unit determines the quantization scale type having a minimum number of assigned bits with respect to the distortion rate of the input signal.

29. The apparatus of claim 26, wherein the quantization scale type is determined from quantization scales having different quantization intervals with respect to a reference point of 0.

30. The apparatus of claim 29, wherein the quantization scale type is defined by using Formula 1:

31. The apparatus of claim 26, wherein the scale type and factor determination unit determines the scale factor by using a standard deviation of the input signal.

32. The apparatus of claim 25, wherein the quantizer calculates a normalization coefficient according to the determined quantization scale type and quantizes the input signal by using the calculated normalization coefficient and the scale factor.

33. An apparatus to encode an input signal, comprising: a type determination unit to determine a quantization scale type indicating a scale type for quantization according to a distribution feature of an input signal;a quantizer to quantize the input signal according to the determined quantization scale type and the distribution feature; andan encoder to encode the quantized input signal.

34. The apparatus of claim 33, wherein the type determination unit comprises: a variable determination unit to determine a generalized Gaussian distribution variable at which a generalized Gaussian distribution matches a distribution of the input signal;a correlation determination unit to determine correlation between a number of assigned bits used to encode the input signal and a distortion rate of the input signal by using the determined generalized Gaussian distribution variable; anda scale type and factor determination unit to determine a quantization scale type corresponding to the distortion rate of the input signal and to determine a scale factor used for quantization from the determined correlation.

35. An apparatus to dequantize an input signal, comprising: a dequantizer to dequantize an input signal according to a quantization scale type corresponding to a distortion rate of the input signal and according to a scale factor used for quantization.

36. The apparatus of claim 35, wherein the quantization scale type is determined by using a correlation between the number of assigned bits to encode the input signal and a distortion rate of the input signal.

37. The apparatus of claim 34, wherein the quantization scale type has a minimum number of bits with respect to the distortion rate of the input signal.

38. The apparatus of claim 34, wherein the quantization scale type is determined from quantization scales having different quantization intervals with respect to a reference point of 0.

39. The apparatus of claim 38, wherein the quantization scale type is defined by using Formula 1:

40. The apparatus of claim 34, wherein the scale factor is determined by using a standard deviation of the input signal.

41. An apparatus to decode an input signal, comprising: a decoder to decode an input signal; anda dequantizer to dequantize the decoded input signal according to a quantization scale type corresponding to a distortion rate of the decoded input signal and according to a scale factor used for quantization.

42. The apparatus of claim 41, wherein the quantization scale type comprises quantization scales having different quantization intervals with respect to a reference point of 0.

43. A method of encoding an audio signal, comprising: receiving an audio signal including PCM (pulse code modulation) data;frequency-transforming the audio signal;determining a quantization scale type by analyzing a distribution feature of the frequency-transformed audio signal;quantizing the frequency-transformed audio signal according to the determined quantization scale type;lossless-encoding the quantized signal; andgenerating a bit stream of the lossless-encoded signal.

44. An apparatus to encode an audio signal, comprising: a data input unit to receive an audio signal including PCM (pulse code modulation data);a frequency transforming unit to frequency-transform the audio signal;a scale type determination unit to determine a quantization scale type by analyzing a distribution feature of the frequency-transformed signal;a quantizer to quantize the frequency-transformed audio signal according to the determined quantization scale type;an encoder to lossless-encode the quantized signal; anda bit stream generator to generate a bit stream of the lossless-encoded signal.

45. A method of decoding an audio signal, comprising: receiving an encoded audio bit stream;lossless-decoding the input audio bit stream;dequantizing the decoded audio data according to a quantization scale type corresponding to a distortion rate of the input signal and a scale factor for quantization; andinverse-transforming the dequantized audio data into audio data of a time domain.

46. An apparatus to decode an audio signal, comprising: a data input unit to receive an encoded audio bit stream;a decoder to lossless-decode the input audio bit stream;a dequantizer to dequantize the decoded audio data according to a quantiztion scale type corresponding to a distortion rate of the input signal and according to a scale factor used for quantization; andan inverse-transforming unit to inverse-transform the dequantized audio data into audio data of a time domain.

47. A method of encoding an input signal, the method comprising: encoding an input signal according to a quantization scale type and a scale factor corresponding to a distribution feature of the input signal; andgenerating the encoded signal information on the quantization scale type and the scale factor corresponding to the distribution feature.

48. An apparatus to encode an input signal, comprising: an encoder to encode an input signal according to a quantization scale type and a scale factor corresponding to a distribution feature of the input signal, and to generate the encoded signal and information on the quantization scale type and the scale factor corresponding to the distribution feature.

49. A method of decoding an input signal, the method comprising: decoding an input signal indicating according to a quantization scale type according to a distribution feature of the input signal; anda processor to process the input signal according to the determined quantization scale type and the distribution feature.

50. An apparatus to decode an input signal, comprising: a decoder to decode an input signal to extract a quantization scale type and a scale factor, and to quantize the decoded signal according to the extracted quantization scale type and the scale factor.

51. A method of processing an input signal, the method comprising: determining a quantization scale type according to a distribution feature of the input signal, the distribution feature of the input signal comprising a generalized Gaussian distribution variable which represents a correlation between a number of assigned bits and a distortion rate of the input signal; andprocessing the input signal according to the determined quantization scale type and the distribution feature.

52. The processing method of claim 51, wherein determining the quantization scale type further comprises: determining the generalized Gaussian distribution variable to be a generalized Gaussian distribution variable at which a chi-square scale becomes a minimum value and at which the actual distribution of the input signal becomes similar to the generalized Gaussian distribution variable.

53. The processing method of claim 52, wherein the chi-square scale is expressed as:

54. An apparatus to process an input signal, comprising: a scale type determination unit to determine a quantization scale type according to a distribution feature of the input signal, the distribution feature of the input signal comprising a generalized Gaussian distribution variable which represents a correlation between a number of assigned bits and a distortion rate of the input signal; anda processor to process the input signal according to the determined quantization scale type and the distribution feature.

55. The input signal processing apparatus of claim 54, wherein the scale type determination unit also determines a scale factor according to a distribution feature of the input signal which has a minimum number of assigned bits with respect to the distortion rate of the input signal.

56. An apparatus to encode and decode an input signal, comprising: an encoder to encode an input signal according to a quantization scale type and a scale factor corresponding to a distribution feature of the input signal, and to generate the encoded signal and information on the quantization scale type and the scale factor corresponding to the distribution feature; anda decoder to decode the encoded input signal to extract the quantization scale type and the scale factor, and to quantize the decoded signal according to the extracted quantization scale type and the scale factor.