Apparatus and methods for recompression of digital images

• US 10,362,309 B2
• Filed: 12/04/2017
• Issued: 07/23/2019
• Est. Priority Date: 10/05/2009
• Status: Active Grant

• Current Assignee: Beamr Imaging Ltd.
• Original Assignee: Beamr Imaging Ltd.
• Inventors: Carmel, Sharon, Gill, Dror, Shoham, Tamar
• Primary Examiner(s): Tsai, Tsung Yin
• Application Number:
  US15/830,190
• Publication Number:
  US 20180091814A1
• Time in Patent Office: 596 Days
• 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 19/136 Incoming video signal characteristics or properties
• CPC Class Codes: H04N 19/14 Coding unit complexity, e.g. amount of activity or edge presence estimation H04N19/146 takes precedence
• 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 19/17 the unit being an image region, e.g. an object
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/182 the unit being a pixel
• CPC Class Codes: H04N 19/192 the adaptation method, adaptation tool or adaptation type being iterative or recursive
• CPC Class Codes: H04N 19/50 using predictive coding H04N19/61 takes precedence
• CPC Class Codes: H04N 19/60 using transform coding
• CPC Class Codes: H04N 19/625 using discrete cosine transform [DCT]
• CPC Class Codes: H04N 19/86 involving reduction of coding artifacts, e.g. of blockiness

Abstract

A system and method for generating a second reduced size digital image from a first digital image, the method including iteratively compressing the first digital image to an extent determined by a quality measure comprising at least a blockiness measure quantifying added artifactual edges along coding block bomldaries of the second image and/or use of a quantization matrix generated by computing a weighted average of the quantization matrix of the first digital image and a scaled second quantization matrix.

Claims

1. A computerized method of iteratively searching for a target value of a compression parameter to be used in optimized image compression of an input image, the method comprising:

selecting an initial value of the compression parameter at least based on one or more characteristics of the input image;

compressing the input image using the initial value, giving rise to a candidate compressed image;

determining whether quality of the candidate compressed image meets a perceptual quality criterion; and

in response to the perceptual quality criterion being not met, iteratively searching for a target value of the compression parameter by;

iteratively, generating an updated value of the compression parameter and repeating said compressing and determining using the updated value, until the perceptual quality criterion is met by an updated compressed image, and selecting the updated value used to obtain the updated compressed image to be the target value of the compression parameter.

2. The computerized method of claim 1, wherein the compression parameter is a Modified Quality Factor (MQF), and said compressing in each iteration is performed using a quantization matrix derived at least from a scaled quantization matrix, the scaled quantization matrix being obtained by scaling a default quantization matrix according to a factor derived from the MQF.

3. The computerized method of claim 1, further comprising:

selecting a step size to be used in the iterative searching, wherein the step size is selected based on at least one of the one or more characteristics of the input image.

4. The computerized method of claim 1, wherein the input image is a compressed input image, and the compressing comprises a process of recompression.

5. The computerized method of claim 4, wherein the compressed input image has an input quantization matrix associated therewith, and the one or more characteristics of the compressed input image comprise the input quantization matrix.

6. The computerized method of claim 4, wherein the one or more characteristics of the compressed input image comprise a characteristic derived from an analysis of dequantized DCT coefficients associated with the compressed input image.

7. The computerized method of claim 1, wherein the iterative searching of a target value employs one or more of the following approaches:

a bi-section approach, a modified bi-section approach and a secant approach.

8. The computerized method of claim 3, wherein at least one of the selecting an initial value and selecting a step size is based on one or more previous target values of the compression parameter selected for compressing one or more previous images having similar characteristics as the one or more characteristics of the input image.

9. The computerized method of claim 1, wherein the input image is divided into a plurality of regions, and wherein said selecting an initial value, compressing, determining and iteratively searching of a target value are performed iteratively for each region of the plurality of regions, giving rise to a plurality of target values of the compression parameter corresponding to the plurality of regions.

10. The computerized method of claim 9, wherein the selecting an initial value for a subsequent region of the plurality of regions is based on one or more previous target values selected for compressing one or more regions preceding the subsequent region.

11. The computerized method of claim 9, further comprising selecting a step size to be used in the iteratively searching of a target value for each given region of the plurality of regions, wherein the step size to be used for a subsequent region is based on a distance between one or more previous target values selected for compressing one or more regions preceding the subsequent region and one or more corresponding initial values.

12. The computerized method of claim 9, further comprising performing the optimized image compression of the input image by compressing the input image using the plurality of target values.

13. A computerized system of iteratively searching for a target compression parameter to be used in optimized image compression of an input image, the system comprising a processor operatively coupled to a memory, the processor configured to:

select an initial value of the compression parameter at least based on one or more characteristics of the input image;

compress the input image using the initial value, giving rise to a candidate compressed image;

determine whether quality of the candidate compressed image meets a perceptual quality criterion; and

in response to the perceptual quality criterion being not met, iteratively search for a target value of the compression parameter by;

iteratively, generating an updated value of the compression parameter and repeating said compressing and determining using the updated value, until the perceptual quality criterion is met by an updated compressed image, and selecting the updated value used to obtain the updated compressed image to be the target value of the compression parameter.

14. The computerized system of claim 13, wherein the compression parameter is a Modified Quality Factor (MQF), and the processor is configured to compress the input image in each iteration using a quantization matrix derived at least from a scaled quantization matrix, the scaled quantization matrix being obtained by scaling a default quantization matrix according to a factor derived from the MQF.

15. The computerized system of claim 13, wherein the processor is further configured to select a step size to be used in the iterative searching, and wherein the step size is selected based on at least one of the one or more characteristics of the input image.

16. The computerized system of claim 13, wherein the input image is a compressed input image, and the compressing comprises a process of recompression.

17. The computerized system of claim 16, wherein the compressed input image has an input quantization matrix associated therewith, and the one or more characteristics of the compressed input image comprise the input quantization matrix.

18. The computerized system of claim 15, wherein the processor is configured to perform at least one of the selecting an initial value and selecting a step size based on one or more previous target values of the compression parameter selected for compressing one or more previous images having similar characteristics as the one or more characteristics of the input image.

19. The computerized system of claim 13, wherein the input image is divided into a plurality of regions, and wherein the processor is configured to perform said selecting an initial value, compressing, determining and iteratively searching of a target value iteratively for each region of the plurality of regions, giving rise to a plurality of target values of the compression parameter corresponding to the plurality of regions.

20. The computerized system of claim 19, wherein the processor is configured to perform the selecting an initial value for a subsequent region of the plurality of regions based on one or more previous target values selected for compressing one or more regions preceding the subsequent region.

21. The computerized system of claim 19, wherein the processor is further configured to perform the optimized image compression of the input image by compressing the input image using the plurality of target values.

22. A non-transitory computer readable storage medium tangibly embodying a program of instructions executable by a computer to perform a method of iteratively searching for a target value of a compression parameter to be used in optimized image compression of an input image, the method comprising:

selecting an initial value of the compression parameter at least based on one or more characteristics of the input image;

compressing the input image using the initial value, giving rise to a candidate compressed image;

determining whether quality of the candidate compressed image meets a perceptual quality criterion; and

in response to the perceptual quality criterion being not met, iteratively searching for a target value of the compression parameter by;

iteratively, generating an updated value of the compression parameter and repeating said compressing and determining using the updated value, until the perceptual quality criterion is met by an updated compressed image, and selecting the updated value used to obtain the updated compressed image to be the target value of the compression parameter.