Distortion-adaptive visual frequency weighting

• US 7,035,473 B1
• Filed: 12/08/2000
• Issued: 04/25/2006
• Est. Priority Date: 03/01/2000
• Status: Expired due to Fees

• Current Assignee: Sharp Kabushiki Kaisha (Hon Hai Precision Industry Co., Ltd.)
• Original Assignee: Sharp Laboratories of America Incorporated (Hon Hai Precision Industry Co., Ltd.)
• Inventors: Zeng, Wenjun, Lei, Shawmin
• Primary Examiner(s): Mancuso, Joseph
• Assistant Examiner(s): SHERALI, ISHRAT I
• Application Number:
  US09/687,727
• Time in Patent Office: 1,964 Days
• Field of Search: 382232-253, 37524002-24003, 37524011-2402, 375/240.22, 35842601-42602
• US Class Current: 382/251
• 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 19/154 Measured or subjectively estimated visual quality after decoding, e.g. measurement of distortion use of rate-distortion criteria H04N19/147
• CPC Class Codes: H04N 19/184 the unit being bits, e.g. of the compressed video stream
• CPC Class Codes: H04N 19/60 using transform coding
• CPC Class Codes: H04N 19/63 using sub-band based transform, e.g. wavelets
• CPC Class Codes: H04N 19/64 characterised by ordering of coefficients or of bits for transmission

Abstract

The ability of the visual system to detect contrast in an image is a function of the frequency of the contrasting pattern and the distortion of the image. The visual system is more sensitive to contrasting patterns of lower frequency. When the image is significantly distorted, the visual system is even more sensitive to lower frequencies than higher frequencies. An image encoder employs lossy data compression processes producing a distorted reconstructed image. A method of quantizing image data including the step of varying the magnitude of a quantization step as a function of the distortion of an image is disclosed for further visually optimizing image quantization. Another method utilizes distortion adaptive weighting to vary the limit of code block truncation during embedded bitstream coding to visually optimize image compression by increasing relative lossiness of compression at higher frequencies.

Claims

1. A method of compressing an image comprising the steps of:

(a) separating data representing said image into a plurality of image data frequency sub-bands;

(b) transforming said data to a plurality of transform coefficients;

(c) mapping said transform coefficients to a plurality of quantizer indices, each said quantizer index comprising a plurality of digits arrayed from a most significant digit to a least significant digit;

(d) adding said most significant digits of said quantizer indices representing an image data frequency sub-band to a bitstream;

(e) repeating step (d) for a less significant digit of said quantizer indices until a number of significant digits specified by a truncation limit for said image data frequency sub-band is reached; and

(f) varying said truncation limit for at least two of said image data frequency sub-bands as a function of a distortion of said image, wherein the magnitude of the quantization is increased for higher frequencies relative to lower frequencies as the data rate of transmission of said image data decreases, wherein the magnitude of the quantization is decreased for higher frequencies relative to lower frequencies as the data rate of transmission of said image data increases.

2. The method of claim 1 further comprising the step of varying said truncation limit as a function of a frequency of said image data represented by said image data frequency sub-band.

3. The method of claim 1 wherein the step of varying said truncation limit for at least two of said image data frequency sub-bands as a function of a distortion of said image comprises varying said truncation limit to increase a number of significant digits added to said bit stream for a lower frequency image sub-band relative to a number of significant digits added to said bit stream for a higher frequency sub-band as said distortion of said image increases.

4. The method of claim 3 further comprising the step of varying said truncation limit as a function of a frequency of said image data represented by said image data frequency sub-band.