Process for adaptive quantization for the purpose of data reduction in the transmission of digital images

Claim 2

A process for adaptive quantization for the purpose of data reduction in the transmission of digital images, comprising the steps of: a) subdividing an image region to be transmitted into a multiplicity of blocks; b) subdividing each block into subregions E(1), where l=1, . . . , L; c) determining an activity measure A(1) for each subregion E(1); d) determining a subregion parameter P(1) for each subregion E(1) from the activity measure A(1) using a function Q O in accordance with the relationship P(1)=Q O (A(1)), e) calculating a block parameter P by summing over all the subregions E(1) of each block only after using the function Q O for each subregion E(1), the block parameter being calculated in accordance with the relationship ##EQU3## and being characteristic of each block, f) determining a quantization parameter Q=(F.sub.Q /L) . P for each block from the block parameter P and using a scaling factor F Q , g) subjecting the image region to be transmitted in blockwise fashion to a transformation, transformation coefficients being determined which are quantized for each block as a function of the value of the quantization parameter Q belonging to the respective block, and h) transmitting the transformed image region.

Claim 3

The process as claimed in claim 1, wherein the activity measure is given by the relationship A=Σ(/y(m,n+1)-y(m,n) /+/y(m+1,n)-y(m,n) /) m,n ε E(1) where m and n respectively are indices of pixels within a subregion of a block of the image region, and y is an image signal of the image region.

Claim 4

The process as claimed in claim 1, wherein discrete cosine transformation is used to transform the image regions to be transmitted.

Claim 5

The process as claimed in claim 1, wherein each of the blocks has a plurality of pixels arranged in columns and rows and wherein each of the subregions of the blocks comprise a pixel and its nearest adjacent pixel in a subsequent row and its nearest adjacent pixel in a subsequent column.

Claim 6

A process for adaptive quantization for the purpose of data reduction in the transmission of digital images, comprising the steps of: a) providing an image region to be transmitted and subdividing the image region to be transmitted into a multiplicity of blocks; b) subdividing each block into subregions E(1), where l=1, . . . , L; c) determining an activity measure A(1) for each subregion E(1); d) determining a subregion parameter P(1) for each subregion E(1) from the activity measure A(1) using a function Q O in accordance with the relationship P(1)=Q O (A(1)), e) calculating a block parameter P by summing over all the subregions E(1) of each block only after using the function Q O for each subregion E(1), the block parameter being calculated in accordance with the relationship ##EQU4## and being characteristic of each block, f) determining a quantization parameter Q=(F.sub.Q /L) . P for each block from the block parameter P and using a scaling factor F Q , g) subjecting the image region to be transmitted in blockwise fashion to a transformation, transformation coefficients being determined which are quantized for each block as a function of the value of the quantization parameter Q belonging to the respective block, wherein the quantized transformation coefficients represent a reduction of data relative to the image region to be transmitted, and h) transmitting the image region with the reduced data.

Claim 7

A process for adaptive quantization for the purpose of data reduction in the transmission of digital images, comprising the steps of: a) providing an image region to be transmitted and subdividing the image region to be transmitted into a multiplicity of blocks; b) subdividing each block into subregions E(1), where l=1, . . . , L; c) determining an activity measure A(1) for each subregion E(1); d) determining a subregion parameter P(1) for each subregion E(1) from the activity measure A(1) using a function Q O in accordance with the relationship P(1)=Q O (A(1)), e) calculating a block parameter P by summing over all the subregions E(1) of each block only after using the function Q O for each subregion E(1), the block parameter being calculated in accordance with the relationship ##EQU4## and being characteristic of each block, f) determining a quantization parameter Q=(F.sub.Q /L) . P for each block from the block parameter P and using a scaling factor F Q , g) subjecting the image region to be transmitted in blockwise fashion to a transformation, transformation coefficients being determined which are quantized for each block as a function of the value of the quantization parameter Q belonging to the respective block, wherein the quantized transformation coefficients represent a reduction of data relative to the image region to be transmitted, and h) transmitting the image region with the reduced data.

Claim 8

The process as claimed in claim 5, wherein the activity measure is given by the relationship A=Σ(/y(m,n+1)-y(m,n) /+/y(m+1,n)-y(m,n)/) m,n ε E(1) where m and n respectively are indices of pixels within a subregion of a block of the image region, and y is an image signal of the image region.

Claim 9

The process as claimed in claim 5, wherein discrete cosine transformation is used to transform the image regions to be transmitted.

Claim 10

The process as claimed in claim 5, wherein each of the blocks has a plurality of pixels arranged in columns and rows and wherein each of the subregions of the blocks comprise a pixel and its nearest adjacent pixel in a subsequent row and its nearest adjacent pixel in a subsequent column.