Method of compressing digital images
First Claim
1. A method of compressing a digital representation of an image into a one-dimensional bitstream, the digital representation comprising a two-dimensional array of pixels wherein a primary color component and secondary color components are associated with each pixel, the method comprising:
- representing the image in a series of quality scales of progressively decreasing quality, wherein a higher quality scale comprises more data than a lower quality scale, and wherein lower quality scales are formed from higher quality scales by decreasing the number of stored color components or by decreasing the number of pixels;
representing the image in terms of quality scales in terms of a base quality scale image and differential images wherein a differential image at a given quality scale is the difference between the image at the given quality scale and a representation of the image scaled up from a reconstructed representation at the adjacent lower quality scale, the reconstructed representation determined by a process comprising;
transforming the image into a set of coefficients associated with known functions, quantizing the set of coefficients by dividing by quantization values and rounding to integer values, dequantizing the set of quantized coefficients by multiplying by the quantization values, and performing the inverse transform associated with the known functions to produce a reconstructed representation;
representing the base quality scale image and the differential images as integer values by a process comprising transforming to a set of coefficients associated with known functions and quantizing the set of coefficients by dividing by quantization values and rounding to integer values; and
encoding the integer values corresponding to the lowest quality scale and the differential images by a lossless ordered statistics encoding process to produce a one-dimensional bitstream.
1 Assignment
0 Petitions
Accused Products
Abstract
A method of compressing digital representations of images stores the images in multiple subsampling quality scales in a progressive manner such that a higher quality scale contains only data incremental to the data in an adjacent lower quality scale. The data in each quality scale is transformed, quantized, and entropy encoded. A discrete wavelet cosine transformation combining frequency transformation features of a discrete cosine transformation and spatial transformation features of a wavelet transformation is provided. Sequences of subsampling quality scales are provided for multi-scale representations of images. A novel context selection method is used which treats color components and coefficients of different positions differently. An image compressed in a given quality scale can be transmitted or decompressed progressively allowing progressive downloading or transmission over the Internet.
-
Citations
31 Claims
-
1. A method of compressing a digital representation of an image into a one-dimensional bitstream, the digital representation comprising a two-dimensional array of pixels wherein a primary color component and secondary color components are associated with each pixel, the method comprising:
-
representing the image in a series of quality scales of progressively decreasing quality, wherein a higher quality scale comprises more data than a lower quality scale, and wherein lower quality scales are formed from higher quality scales by decreasing the number of stored color components or by decreasing the number of pixels;
representing the image in terms of quality scales in terms of a base quality scale image and differential images wherein a differential image at a given quality scale is the difference between the image at the given quality scale and a representation of the image scaled up from a reconstructed representation at the adjacent lower quality scale, the reconstructed representation determined by a process comprising;
transforming the image into a set of coefficients associated with known functions, quantizing the set of coefficients by dividing by quantization values and rounding to integer values, dequantizing the set of quantized coefficients by multiplying by the quantization values, and performing the inverse transform associated with the known functions to produce a reconstructed representation;
representing the base quality scale image and the differential images as integer values by a process comprising transforming to a set of coefficients associated with known functions and quantizing the set of coefficients by dividing by quantization values and rounding to integer values; and
encoding the integer values corresponding to the lowest quality scale and the differential images by a lossless ordered statistics encoding process to produce a one-dimensional bitstream. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 26, 27, 28, 29, 30, 31)
recovering context prediction probability tables used in compression;
decoding integer values corresponding to the base quality scale representation and the differential images;
reverse ordering each decoded integer value to a two-dimensional position in an array of pixels;
multiplying each decoded integer value corresponding to the base quality scale and the differential images by the quantization value;
performing the inverse transform associated with the known functions to reconstruct the digital representation of the base quality scale and the differential images;
upscaling the digital representation of an image at a lower quality scale to the next higher quality scale; and
adding a differential image at a given quality scale to the upscaled image at the given quality scale to reconstruct the digital representation at the given quality scale.
-
-
16. The method of claim 15 wherein for a primary color component, a context used in context prediction tables comprises a positional index and neighboring coefficients of primary color pixels, for a first secondary color component, a context comprises a positional index, coefficients of neighboring first secondary color components, and the coefficient of the primary color component of the same positional index, and for a second secondary color component, a context comprises a positional index, neighboring second secondary color coefficients, and the coefficients of the primary and first secondary color components of the same positional index.
-
17. The method of claim 16 wherein context prediction tables are associated with quantized coefficients considered in groups, wherein a first group comprises the coefficient corresponding to a lowest positional index, a second group comprises a first row of coefficients excluding the first group coefficient, a third group comprises a first column of coefficients excluding the first group coefficient, and a fourth group comprises the remaining coefficients, and wherein the context is different for each group.
-
18. The method of claim 15 wherein the inverse transform associated with the known functions is the inverse of a transform combining frequency transform features and multi-resolution transform features.
-
19. The method of claim 18 wherein the inverse transform is defined recursively from an inverse discrete cosine transform and a permutation function, and wherein the input elements of the inverse transform are separated into a first portion and a second portion, the first portion containing lower scales of representation of input to the transform.
-
20. The method of claim 19 wherein the inverse transform is an inverse discrete wavelet transform.
-
26. An apparatus comprising instructions for reconstructing a digital representation of an image compressed by the method of claim 1, the instructions comprising instructions for:
-
recovering context prediction probability tables used in compression;
decoding integer values corresponding to the base quality scale representation and the differential images;
reverse ordering each decoded integer value to a two-dimensional position in an array of pixels;
multiplying each decoded integer value corresponding to the base quality scale and the differential images by the quantization value;
performing the inverse transform associated with the known functions to reconstruct the digital representation of the base quality scale and the differential images;
upscaling the digital representation of an image at a lower quality scale to the next higher quality scale; and
adding a differential image at a given quality scale to the upscaled image at the given quality scale to reconstruct the digital representation at the given quality scale.
-
-
27. The apparatus of claim 26 wherein for the primary color component, the context used in context prediction tables comprises a positional index and neighboring coefficients of primary color pixels, for a first secondary color component, the context comprises a positional index, coefficients of neighboring first secondary color components, and the coefficient of the primary color component of the same positional index, and for a second secondary color component, the context comprises a positional index, neighboring second secondary color coefficients, and the coefficients of the primary and first secondary color components of the same positional index.
-
28. The apparatus of claim 27 wherein context prediction tables are associated with quantized coefficients considered in groups, wherein a first group comprises the coefficient corresponding to a lowest positional index, a second group comprises a first row of coefficients excluding the first group coefficient, a third group comprises a first column of coefficients excluding the first group coefficient, and a fourth group comprises the remaining coefficients, and wherein the context is different for each group.
-
29. The apparatus of claim 26 wherein the inverse transform associated with the known functions is the inverse of a transform combining frequency transform features and multi-resolution transform features.
-
30. The apparatus of claim 29 wherein the inverse transform is defined recursively from an inverse discrete cosine transform and a permutation function, and wherein the input elements of the inverse transform are separated into a first portion and a second portion, the first portion containing lower scales of representation of input to the transform.
-
31. The apparatus of claim 30 wherein the inverse transform is an inverse discrete wavelet transform.
-
21. An apparatus comprising instructions for performing a method of compressing a digital representation of an image into a one-dimensional bitstream, the digital representation comprising a two-dimensional array of pixels wherein a primary color component and secondary color components are associated with each pixel, the method comprising:
-
representing the image in a series of quality scales of progressively decreasing quality, wherein a higher quality scale comprises more data than a lower quality scale, and wherein lower quality scales are formed from higher quality scales by decreasing the number of stored color components or by decreasing the number of pixels;
representing the image in terms of quality scales in terms of a base quality scale image and differential images wherein a differential image at a given quality scale is the difference between the image at the given quality scale and a representation of the image scaled up from a reconstructed representation at the adjacent lower quality scale, the reconstructed representation determined by a process comprising;
transforming the image into a set of coefficients associated with known functions, quantizing the set of coefficients by dividing by quantization values and rounding to integer values, dequantizing the set of quantized coefficients by multiplying by the quantization values, and performing the inverse transform associated with the known functions to produce a reconstructed representation;
representing the base quality scale image and the differential images as integer values by a process comprising transforming to a set of coefficients associated with known functions and quantizing the set of coefficients by dividing by quantization values and rounding to integer values; and
encoding the integer values corresponding to the lowest quality scale representation and the differential images by a lossless ordered statistics encoding process to produce a one-dimensional bitstream. - View Dependent Claims (22, 23, 24, 25)
-
Specification