Quantization table adjustment

• US 6,023,531 A
• Filed: 06/24/1997
• Issued: 02/08/2000
• Est. Priority Date: 12/13/1991
• Status: Expired due to Fees

• Current Assignee: Avid Technology Incorporated
• Original Assignee: Avid Technology Incorporated
• Inventors: Peters, Eric C.
• Primary Examiner(s): Couso, Jose L.
• Application Number:
  US08/881,754
• Time in Patent Office: 959 Days
• Field of Search: 382/234, 382/254, 382/250, 382/251, 382/276, 382/279, 348/403, 348/405, 358/432, 358/433
• US Class Current: 382/232
• CPC Class Codes: H04N 19/117 Filters, e.g. for pre-processing or post-processing sub-band filter banks H04N19/635
• CPC Class Codes: H04N 19/124 Quantisation
• CPC Class Codes: H04N 19/126 Details of normalisation or weighting functions, e.g. normalisation matrices or variable uniform quantisers
• CPC Class Codes: H04N 19/15 by monitoring actual compressed data size at the memory before deciding storage at the transmission buffer
• CPC Class Codes: H04N 19/154 Measured or subjectively estimated visual quality after decoding, e.g. measurement of distortion use of rate-distortion criteria H04N19/147
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/18 the unit being a set of transform coefficients
• CPC Class Codes: H04N 19/60 using transform coding

Abstract

The method for adjusting quality during image capture includes computing a discrete cosine transform of a digital image to create DCT coefficients. A quantization table is generated that specifies frequency bands to be filtered and the DCT coefficients are digitized using the quantization table. It is preferred that the DCT coefficients be ordered in a zig-zag sequence to facilitate run-length encoding.

Claims

1. A method for compressing digitized signals digitized by a digitizer, comprising the steps of:

computing a discrete cosine transform of a first plurality of the signals to create a first plurality of coefficients corresponding to frequencies;

specifying a cutoff frequency above which the first plurality of the computed coefficients are to be eliminated;

creating a first quantization table including a first plurality of table values indicative of the cutoff frequency, each table value corresponding to one of a plurality of frequencies;

quantizing the first plurality of computed coefficients using the first quantization table and outputting compressed signals;

computing a discrete cosine transform of a second plurality of the signals to create a second plurality of coefficients corresponding to the frequencies;

adjusting the cutoff frequency above which the second plurality of the computed coefficients are to be eliminated;

creating a second quantization table including a second plurality of table values indicative of the cutoff frequency after the step of adjusting, wherein each of the second plurality of table values corresponds to one of the plurality of frequencies, and wherein at least some of the values in the first table that correspond respectively to a subset of the frequencies are different from at least some of the values in the second table that correspond respectively to the same subset of the frequencies;

quantizing the second plurality of computed coefficients using the second table and outputting further compressed signals; and

decompressing the compressed signals that were compressed according to the first table and decompressing the further compressed signals that were compressed according to the second table using a same table.

2. The method of claim 1, wherein the step of creating includes setting the table values below the cutoff frequency to predetermined values, to mute the effects of eliminating the computed coefficients above the cutoff frequency.

3. The method of claim 1 wherein the steps of computing and the steps of quantizing operate in accordance with the JEPG standard.

4. An apparatus for compressing digitized image signals digitized by a digitizer, comprising:

means for receiving the signals, for computing an orthogonal transform of the signals and for outputting a plurality of coefficients corresponding to frequencies;

means for creating a quantization table including a plurality of table values indicative of a cutoff frequency above which the computed coefficients are to be eliminated, each table value corresponding to a frequency;

means for adjusting the cutoff frequency to create a quantization table including a different plurality of table values indicative of a different frequency above which computed coefficients are to be eliminated;

means for receiving the plurality of coefficients and for quantizing the plurality of coefficients using the quantization table values; and

means for decompressing the image signals processed by the means for receiving and the means for quantizing, the means for decompressing using a sample table independent of adjustment by the means for adjusting to create the different plurality of table values for the images to be decompressed.

5. The apparatus of claim 4 wherein the means for receiving and quantizing operate in accordance with the JEPG standard.

6. An apparatus for compressing signals representative of multidimensional physical data comprising:

an orthogonal transformer which receives the signals, computes an orthogonal transform of the signals and outputs a plurality of transform coefficients corresponding to frequencies; and

a quantizer including;

an input for receiving the plurality of computed coefficients;

a quantization table comprising a plurality of table values including table values indicative of a cutoff frequency above which the computed coefficients are to be eliminated, and each table value corresponding to a frequency;

means for adjusting the quantization table to be indicative of an adjusted cutoff frequency, wherein the step of adjusting the quantization table changes values in the table that correspond to at least one of the frequencies; and

a decompressor for decompressing the image signals compressed by the transformer and the quantizer using a same table independent of tables value changes due to adjustment by the means for adjusting.

7. The apparatus of claim 6 wherein the orthogonal transformer operates in accordance with the JEPG standard.

8. A method for compressing signals representative of a series of two-dimensional images, comprising:

receiving signals representative of a first of the images,

transforming the signals of the first of the images to obtain lower frequency coefficients below a cut-off frequency and higher frequency coefficients above the cut-off frequency,

quantizing the coefficients obtained in the step of transforming, the step of quantizing including the steps of;

reducing the precision of the lower frequency coefficients,

reducing the higher frequency coefficients to zero,

adjusting the cut-off frequency,

receiving signals representative of a second of the images,

transforming the signals representative of the second of the images to obtain further lower frequency coefficients below the adjusted cut-off frequency and further higher frequency coefficients above the cut-off frequency,

quantizing the coefficients obtained in the step of transforming the signals representative of the second of the images, the step of quantizing including the steps of;

reducing the precision of the further lower frequency coefficients,

reducing the further higher frequency coefficients to zero, wherein the step of reducing the higher frequency coefficients to zero and the step of reducing the further higher frequency coefficients to zero operate on different numbers of coefficients, and wherein the step of reducing the precision of the lower frequency coefficients and the step of reducing the precision of the further lower frequency coefficients operate on different numbers of coefficients, and

decompressing the first and second images using a same quantization table.

9. The method of claim 8, wherein the steps of quantizing mute the frequency components just below the cut-off frequency.

10. The method of claim 8, further including the step of digitizing the images to obtain the signals representative of the images.

11. The method of claim 10, wherein the steps of transforming and the steps of quantizing operate in accordance with the JEPG standard.

12. The method of claim 11, wherein the steps of quantizing mute the frequency components just below the cut-off frequency.

13. The method of claim 8, further including the step of digitizing the images to obtain the signals representative of the images.

14. The method of claim 8, wherein the steps of transforming and the steps of quantizing operate in accordance with the JEPG standard.

15. Apparatus for compressing signals representative of a series of two-dimensional images, comprising:

means for receiving signals representative of a first and a second of the images,

means for transforming the signals of the first and the second of the images to obtain lower frequency coefficients below a cut-off frequency and higher frequency coefficients above the cut-off frequency,

means for quantizing the coefficients obtained from the means for transforming, the means for quantizing including;

means for reducing the precision of the lower frequency coefficients, the means for reducing the precision being operative to reduce the precision of different numbers of coefficients based on the cut-off frequency,

means for reducing the higher frequency coefficients to zero, the means for reducing the higher frequency coefficients to zero being operative to reduce different numbers of coefficients to zero based on the cut-off frequency, and

means for adjusting the cut-off frequency between the quantizing of the first and second images, and

means for decompressing the first and second images using a same quantization table, despite changes in the number of coefficients whose precision was reduced and changes in the number of coefficients reduced to zero.

16. The apparatus of claim 15, wherein the means for quantizing is constructed and adapted to mute the frequency components just below the cut-off frequency.

17. The method of claim 15, further including means for digitizing the images to obtain the signals representative of the images.

18. The method of claim 17, wherein the means for transforming and the means for quantizing operate in accordance with the JEPG standard.

19. The apparatus of claim 18, wherein the means for quantizing is constructed and adapted to mute the frequency components just below the cut-off frequency.

20. The method of claim 15, further including means for digitizing the images to obtain the signals representative of the images.

21. The method of claim 15, wherein the means for transforming and the means for quantizing operate in accordance with the JEPG standard.

22. An apparatus for compressing signals representative of a series of two-dimensional images, comprising:

an orthogonal transformer which receives the signals, computes an orthogonal transform of the signals and outputs a plurality of transform coefficients corresponding to frequencies,

a quantizer including a quantization table comprising a plurality of table values including table values indicative of a cutoff frequency above which the transform coefficients are to be reduced to zero,

means for adjusting the quantization table to be indicative of an adjusted cutoff frequency, wherein the means or adjusting the quantization table is operative to change values in the table that correspond to at least one of the frequencies, and

dequantizer constructed and adapted to dequantize the series of images using a same quantization table independent of tables value changes due to adjustment by the means for adjusting between quantization steps for the series of images.

23. The apparatus of claim 22, wherein the quantizer mutes the frequency components just below the cut-off frequency.

24. The apparatus of claim 22, further including a digitizer for digitizing the images to obtain the signals representative of the images.

25. The method of claim 24, wherein the transformer and the quantizer operate in accordance with the JEPG standard.

26. The apparatus of claim 25, wherein the quantizer mutes the frequency components just below the cut-off frequency.

27. The apparatus of claim 22, further including a digitizer for digitizing the images to obtain the signals representative of the images.

28. The method of claim 22, wherein the transformer and the quantizer operate in accordance with the JEPG standard.

29. A method for adjusting an amount of compression of a plurality of digitized image signals digitized by a digitizer, comprising the steps of:

computing an orthogonal transform of a first plurality of the signals to create a first plurality of coefficients corresponding to frequencies;

creating an original quantization table including a plurality of table values indicative of a cutoff frequency, each of the table values corresponding to a frequency;

quantizing the first plurality of computed coefficient using the quantization table and outputting a first plurality of compressed signals;

computing an orthogonal transform of a second plurality of the signals to create a second plurality of coefficients corresponding to the frequencies;

adjusting the cutoff frequency and the original quantization table values to obtain adjusting quantization table values to control the amount of compression of the compressed signals, wherein at least some of the values in the original table that correspond respectively to a subset of the frequencies are different from at least some of the values in the adjusted table that correspond respectively to the same subset of the frequencies;

quantizing the second plurality of computed coefficients using the quantization table after the step of adjusting, and outputting a second plurality of compressed signals; and

decompressing both the first and second pluralities of compressed signals using a same table.

30. The method of claim 29, wherein the digitized image signals are reconstructed from the compressed signals using the original quantization table as the same table.