Image compression and decompression using controlled quality loss

Patent metadata

Field	Value
Publication number	US-12587684-B2
Application number	US-202418643085-A
Country	US
Kind code	B2
Filing date	Apr 23, 2024
Priority date	Feb 17, 2018
Publication date	Mar 24, 2026
Grant date	Mar 24, 2026

Abstract

The loss of image quality during compression is controlled using a sequence of quality control metrics. The sequence of quality control metrics is selected for quantizing transform coefficients within an area of the image based on an error level definition. Candidate bit costs are then determined by quantizing the transform coefficients according to the error level definition or a modified error level and the sequence of quality control metrics. Where the candidate bit cost resulting from using the modified error level is lower than the candidate bit cost resulting from using the error level definition, the transform coefficients are quantized according to the modified error level and the sequence of quality control metrics. Otherwise, the transform coefficients are quantized based on the error level definition and according to the sequence of quality control metrics.

First claim

What is claimed is:

1 . A method for compressing an image, the method comprising:
selecting, for quantizing transform coefficients within a local area of the image and based on an error level definition representing a maximum quantization error for compressing the transform coefficients, a sequence of quality control metrics defining an order for applying multiple quality control metrics of the sequence of quality control metrics; iteratively processing a first subset of the transform coefficients located within a first portion of the local area of the image according to the error level definition and the sequence of quality control metrics to produce first quantized transform coefficients; iteratively processing a second subset of the transform coefficients located within a second portion of the local area of the image according to a modified error level and the sequence of quality control metrics to produce second quantized transform coefficients, wherein the modified error level represents a quantization error different than the maximum quantization error; and producing a compressed image based on the first quantized transform coefficients and the second quantized transform coefficients.

2 . The method of claim 1 , wherein selecting the sequence of quality control metrics defining the order for applying the multiple quality control metrics of the sequence of quality control metrics comprises:
determining that a subset of quality control metrics available for compressing the transform coefficients is associated with error levels exceeding the error level definition; and excluding one or more quality control metrics of the subset of the quality control metrics from the sequence of quality control metrics.

3 . The method of claim 1 , wherein iteratively processing the first subset of the transform coefficients located within the first portion of the local area of the image according to the error level definition and the sequence of quality control metrics to produce the first quantized transform coefficients comprises:
quantizing the first subset of the transform coefficients using a first quality control metric of the sequence of quality control metrics, wherein quantizing the first subset of the transform coefficients using the first quality control metric results in a quantization error; determining that the quantization error exceeds a quality score calibration associated with the first quality control metric; and responsive to determining that the quantization error exceeds the quality score calibration, further quantizing the first subset of the transform coefficients using a second quality control metric of the sequence of quality control metrics.

4 . The method of claim 1 , wherein iteratively processing the second subset of the transform coefficients located within the second portion of the local area of the image according to the modified error level and the sequence of quality control metrics to produce the second quantized transform coefficients comprises:
iteratively processing the second subset of the transform coefficients located within the second area of the image according to the error level definition and the sequence of quality control metrics to produce third quantized transform coefficients; determining whether a bit cost for the second quantized transform coefficients is lower than a bit cost for the third quantized transform coefficients; and responsive to determining that the bit cost for the second quantized transform coefficients is not lower than the bit cost for the third quantized transform coefficients, using the third quantized transform coefficients to produce the compressed image instead of the second quantized transform coefficients.

5 . The method of claim 1 , wherein selecting the sequence of quality control metrics defining the order for applying the multiple quality control metrics of the sequence of quality control metrics comprises:
querying a lookup table based on the error level definition.

6 . The method of claim 1 , wherein the quantization error represented by the modified error level is larger than the maximum quantization error.

7 . A method for compressing an image, the method comprising:
selecting, based on an error level definition representing maximum quantization error for compressing transform coefficients of the image, a sequence of quality control metrics defining an order solving multiple quality control metrics of the sequence of quality control metrics; iteratively processing a first subset of transform coefficients of the image according to the error level definition and the sequence of quality control metrics to produce first quantized transform coefficients; iteratively processing a second subset of transform coefficients of the image according to a modified error level and the sequence of quality control metrics to produce second quantized transform coefficients, wherein the modified error level represents a quantization error different than the maximum quantization error; and producing a compressed image based on the first quantized transform coefficients and the second quantized transform coefficients.

8 . The method of claim 7 , wherein selecting the sequence of quality control metrics defining the order for applying the multiple quality control metrics of the sequence of quality control metrics comprises:
excluding one or more quality control metrics available for compressing the image from the sequence of quality control metrics based on the error level definition.

9 . The method of claim 7 , wherein selecting the sequence of quality control metrics defining the order for applying the multiple quality control metrics of the sequence of quality control metrics comprises:
querying a lookup table based on the error level definition.

10 . The method of claim 7 , wherein iteratively processing the first subset of transform coefficients of the image according to the error level definition and the sequence of quality control metrics to produce the first quantized transform coefficients comprises:
determining whether a quantization error resulting from quantizing the first subset of transform coefficients using a first quality control metric of the sequence of quality control metrics exceeds a calibration; and responsive to determining that the quantization error exceeds the calibration, further quantizing the first subset of transform coefficients using a second quality control metric of the sequence of quality control metrics.

11 . The method of claim 7 , wherein iteratively processing the second subset of transform coefficients of the image according to the modified error level and the sequence of quality control metrics to produce the second quantized transform coefficients comprises:
determining whether a bit cost for the second quantized transform coefficients is lower than a bit cost for third quantized transform coefficients produced by iteratively processing the second subset of transform coefficients according to the error level definition and the sequence of quality control metrics; and responsive to determining that the bit cost for the second quantized transform coefficients is lower than the bit cost for the third quantized transform coefficients, using the third quantized transform coefficients to produce the compressed image instead of the second quantized transform coefficients.

12 . The method of claim 7 , wherein the quantization error represented by the modified error level is larger than the maximum quantization error.

13 . A method for compressing an image, the method comprising:
quantizing first transform coefficients of the i

Assignees

• Google Llc

Inventors

• Alakuijala Jyrki
• Obryk Robert
• Kliuchnikov Evgenii
• Szabadka Zoltan
• Wassenberg Jan
• Alakuijala Minttu
• Vandevenne Lode

Classifications

• H04N19/85

  using pre-processing or post-processing specially adapted for video compression · CPC title

• H04N19/184

  the unit being bits, e.g. of the compressed video stream · CPC title

• H04N19/18

  the unit being a set of transform coefficients · CPC title

• H04N19/124 Primary

  Quantisation · CPC title

• H04N19/15

  by monitoring actual compressed data size at the memory before deciding storage at the transmission buffer · CPC title