Claims Analysis
Total Claims: 24
Claim 1
Independent Claim
164 words
A coding system in which discrete coefficients in a frame of blocks of coefficients are quantized with different numbers of quantization bits per coefficient and like numbers of bits for corresponding coefficients in the blocks of the frame, the system comprising, for assigning the quantization bits to the coefficients; means for ordering the coefficients according to the variance of the coefficients through the frame; means for grouping the coefficients in bit groups according to the number of bits, if any, already assigned thereto and for regrouping the coefficients as each bit is assigned; means for determining, for each of a plurality of quantization bits per block, the reduction in quantization error for the frame of blocks with the assignment of a quantization bit to the coefficient of each bit group having the largest variance; and means for comparing the determined reductions in quantization errors and for assigning the next quantization bit to the coefficient for which the largest reduction in quantization error is associated.
Claim 2
Dependent Claim
164 words
A coding system in which discrete coefficients in a frame of blocks of coefficients are quantized with different numbers of quantization bits per coefficient and like numbers of bits for corresponding coefficients in the blocks of the frame, the system comprising, for assigning the quantization bits to the coefficients; means for ordering the coefficients according to the variance of the coefficients through the frame; means for grouping the coefficients in bit groups according to the number of bits, if any, already assigned thereto and for regrouping the coefficients as each bit is assigned; means for determining, for each of a plurality of quantization bits per block, the reduction in quantization error for the frame of blocks with the assignment of a quantization bit to the coefficient of each bit group having the largest variance; and means for comparing the determined reductions in quantization errors and for assigning the next quantization bit to the coefficient for which the largest reduction in quantization error is associated.
Claim 3
Dependent Claim
22 words
A coding system as claimed in claim 1 further comprising means for performing a two dimensional transform to provide the discrete coefficients.
Claim 4
Dependent Claim
34 words
A coding system as claimed in claim 1 wherein the means for determining the reduction in quantization error computes the product of the variance of a coefficient and a normalized change in quantization error.
Claim 5
Dependent Claim
21 words
A coding system as claimed in claim 3 further comprising a lookup table for storing the normalized changes in quantization error.
Claim 6
Dependent Claim
31 words
A coding system as claimed in claim 4 further comprising means to select a table of the normalized changes in quantization error based on the type of distribution of the coefficients.
Claim 7
Dependent Claim
36 words
A coding system as claimed in claim 3 further comprising means to identify the type of distribution of the coefficients and for then determining the normalized change in quantization error based on that type of distribution.
Claim 8
Dependent Claim
27 words
A coding system as claimed in claim 1 further comprising means for transmitting signals indicative of the variances and the type of distribution with the quantized data.
Claim 9
Dependent Claim
229 words
A coding system for transform image coding in which discrete coefficients in a frame of two dimensional blocks of coefficients are quantized with different numbers of quantization bits per coefficient and like numbers of bits for corresponding coefficients in the blocks of the frame, the system comprising, for assigning the quantization bits to the coefficients; means for performing a two dimensional transform to provide the discrete coefficients; means for ordering the coefficients according to the variance of the coefficients through the frame; means for grouping the coefficients in bit groups according to the number of bits, if any, already assigned thereto and for regrouping the coefficients as each bit is assigned; means for providing normalized changes in quantization error as a function of signal distribution; means for determining, for each of a plurality of quantization bits per block, the reduction in quantization error for the frame of blocks with the assignment of a quantization bit to the coefficient of each bit group having the largest variance by computing the product of the variance of the coefficient and the normalized change in quantization error; means for comparing the determined reductions in quantization error and for assigning the next quantization bit to the coefficient for which the largest reduction in quantization error is associated; and means for quantizing the coefficient of each block of coefficients with the assigned number of bits.
Claim 10
Dependent Claim
229 words
A coding system for transform image coding in which discrete coefficients in a frame of two dimensional blocks of coefficients are quantized with different numbers of quantization bits per coefficient and like numbers of bits for corresponding coefficients in the blocks of the frame, the system comprising, for assigning the quantization bits to the coefficients; means for performing a two dimensional transform to provide the discrete coefficients; means for ordering the coefficients according to the variance of the coefficients through the frame; means for grouping the coefficients in bit groups according to the number of bits, if any, already assigned thereto and for regrouping the coefficients as each bit is assigned; means for providing normalized changes in quantization error as a function of signal distribution; means for determining, for each of a plurality of quantization bits per block, the reduction in quantization error for the frame of blocks with the assignment of a quantization bit to the coefficient of each bit group having the largest variance by computing the product of the variance of the coefficient and the normalized change in quantization error; means for comparing the determined reductions in quantization error and for assigning the next quantization bit to the coefficient for which the largest reduction in quantization error is associated; and means for quantizing the coefficient of each block of coefficients with the assigned number of bits.
Claim 11
Independent Claim
168 words
A method of quantizing discrete coefficients in blocks of coefficients with different numbers of quantization bits per coefficient and like numbers of bits per corresponding coefficients in the blocks, the method comprising: for a frame of a plurality of coefficients, ordering the coefficients according to the variance of the coefficients through the frame; assigning a number of quantization bits to each of a plurality of coefficients in each block and grouping the coefficients in bit groups according to the number of thus assigned bits, the bits being assigned by determining, for each of a plurality of quantization bits per block, the reduction in quantization error for the frame of blocks with assignment of the quantization bit to the coefficient of each bit group having the largest variance, comparing the determined reductions in quantization errors and assigning the next quantization bit to the coefficient for which the largest reduction in quantization error is associated; and quantizing the coefficients of each block of coefficients with the assigned number of bits.
Claim 12
Independent Claim
168 words
A method of quantizing discrete coefficients in blocks of coefficients with different numbers of quantization bits per coefficient and like numbers of bits per corresponding coefficients in the blocks, the method comprising: for a frame of a plurality of coefficients, ordering the coefficients according to the variance of the coefficients through the frame; assigning a number of quantization bits to each of a plurality of coefficients in each block and grouping the coefficients in bit groups according to the number of thus assigned bits, the bits being assigned by determining, for each of a plurality of quantization bits per block, the reduction in quantization error for the frame of blocks with assignment of the quantization bit to the coefficient of each bit group having the largest variance, comparing the determined reductions in quantization errors and assigning the next quantization bit to the coefficient for which the largest reduction in quantization error is associated; and quantizing the coefficients of each block of coefficients with the assigned number of bits.
Claim 13
Dependent Claim
21 words
A method as claimed in claim 9 further comprising the step of performing a two-dimensional transform to provide the discrete coefficients.
Claim 14
Dependent Claim
32 words
A method as claimed in claim 9 wherein the reduction in quantization error is determined by computing the product of the variance of a coefficient and a normalized change in quantization error.
Claim 15
Dependent Claim
20 words
A method as claimed in claim 11 wherein the normalized change in quantization error is retrieved from a lookup table.
Claim 16
Dependent Claim
27 words
A method as claim 12 further comprising the step of providing the type of distribution for selecting the normalized change in quantization error from the lookup table.
Claim 17
Dependent Claim
30 words
A method as claimed in claim 11 further comprising the step of identifying the type of distribution of the coefficients in order to determine the normalized change in quantization error.
Claim 18
Dependent Claim
29 words
A method as claimed in claim 9 further comprising the step of transmitting with the quantized coefficients signals indicative of the variances and type of distribution of the coefficients.
Claim 19
Independent Claim
112 words
A method of assigning a number of quantization bits to each of a plurality of discrete coefficients in blocks of coefficients, the method comprising sequentially assigning the available quantization bits to appropriate coefficients and grouping and regrouping the coefficients in bit groups according to the number of thus assigned bits, the bits being assigned by determining, for each assigned quantization bit, the maximum reduction in quantization error for a frame of blocks of coefficients with assignment of the bit to a predetermined one of the coefficients of each bit group, and assigning the bit to the coefficient, throughout the frame of blocks of coefficients, which provides the greatest reduction in quantization error.
Claim 20
Independent Claim
112 words
A method of assigning a number of quantization bits to each of a plurality of discrete coefficients in blocks of coefficients, the method comprising sequentially assigning the available quantization bits to appropriate coefficients and grouping and regrouping the coefficients in bit groups according to the number of thus assigned bits, the bits being assigned by determining, for each assigned quantization bit, the maximum reduction in quantization error for a frame of blocks of coefficients with assignment of the bit to a predetermined one of the coefficients of each bit group, and assigning the bit to the coefficient, throughout the frame of blocks of coefficients, which provides the greatest reduction in quantization error.
Claim 21
Dependent Claim
36 words
A method as claimed in claim 16 further comprising determining the reduction in quantization error from a normalized change in quantization error which is a function of the number of bits already assigned to each coefficient.
Claim 22
Dependent Claim
27 words
A method as claimed in claim 17 wherein the normalized change in quantization error is determined as a function of the type of distribution of the coefficients.
Claim 23
Independent Claim
48 words
A method as claimed in 18 wherein the change in quantization error is computed as the product of the variance and the normalized change in quantization error for the coefficient having the largest variance within each group of coefficients having a particular number of bits already assigned thereto.
Claim 24
Dependent Claim
25 words
A method as claimed in claim 16 wherein the predetermined one of the coefficients of each bit group is the coefficient having the largest variance.