Procedure for marking binary coded data sets

Claim 2

A procedure for marking binary coded data sets, in particular image data and audio signals, by embedding at least one information unit in a data set, comprising the steps of: generating a discrete position sequence using properties characteristics of said data set and using a key; selecting, on the basis of said discrete position sequence, at least one block of image pixels or audio time signals; transforming said selected block of pixels or time signal from the spatial or time domain into a frequency domain wherein said transformed block has a plurality of frequency coefficients; embedding said information unit into said transformed block at positions which are within a predetermined frequency coefficient range.

Claim 3

A procedure according to claim 1 , wherein said data set specific properties are at least one of content-based features, object features, geometric and format specific features.

Claim 4

A procedure according to claim 1 , wherein said transformation function is a discrete cosine transformation, a wavelet transformation, a Fourier transformation, a Hadamard-Ward transformation or a Z transformation.

Claim 5

A procedure according to claim 1 , wherein said blocks transformed into the frequency domain are quantized using quantization matrices.

Claim 6

A procedure according to claim 4 , wherein said quantization matrices correspond to the quantization step used in the JPEG compression standard.

Claim 7

A procedure according to claim 4 , wherein said information units are embedded into said blocks by using said generated position sequence for selecting the positions of the coefficients.

Claim 8

A procedure according to claim 4 , wherein a requantization and retransformation of the data set into a spatial domain occurs after said embedding of said information units.

Claim 9

A procedure according to claim 1 , wherein said embedding of the information units to be integrated occurs at the selected coefficient positions of said blocks by adjusting the size relations of selected frequency coefficient in order to equal a predetermined size relationship among said selected frequency coefficients.

Claim 10

A procedure according to claim 1 , wherein said block includes the entirety of said data set.

Claim 11

A procedure according to claim 1 , wherein said block is a single image pixel or a single audio sample signal.

Claim 12

A procedure according to claim 1 , wherein said block is a set of image pixels/audio signals collected from the data set using said position sequence.

Claim 13

A procedure according to claim 1 , wherein said block is an arbitrary continuous region of the image or any segment of continuous audio signals.

Claim 14

A procedure for marking binary coded bi-level image data sets by embedding at least one information unit in a data set, comprising the steps of: generating a discrete position sequence using property characteristics of the data set and using a key; selecting, on the basis of said discrete position sequence, a plurality of blocks each containing a plurality of picture elements, each of said picture elements having a value of “1” or “0”; specifying a predetermined relationship between the number of picture elements having a “1” value and the number of picture elements having the “0” value; determining the actual relationship between the number of “1” values and the number of “0” values in said selected block; comparing said determined relationship with said predefined relationship and adjusting the relationship between the numbers of “1” and the numbers of “0” in said block until said determined relationship is equal to said predefined relationship wherein the embedding of an information unit equivalent to a “1” bit encoded in a block b satisfies the following requirement: T 1,min <R(b)<T 1,max with R(b)=N 1 /N N 1 =number of “1” bit pixels in a block b T 1,min ; T 1,max upper and lower thresholds N=total number of pixels in the block.

Claim 15

A procedure for marking binary coded bi-level image data sets by embedding at least one information unit in a data set, comprising the steps of: generating a discrete position sequence using property characteristics of the data set and using a key; selecting, on the basis of said discrete position sequence, a plurality of blocks each containing a plurality of picture elements, each of said picture elements having a value of “1” or “0”; specifying a predetermined relationship between the number of picture elements having a “1” value and the number of picture elements having the “0” value; determining the actual relationship between the number of “1” values and the number of “0” values in said selected block; comparing said determined relationship with said predefined relationship and adjusting the relationship between the numbers of “1” and the numbers of “0” in said block until said determined relationship is equal to said predefined relationship wherein the embedding of an information unit equivalent to a “1” bit encoded in a block b satisfies the following requirement: T 1,min <R(b)<T 1,max with R(b)=N 1 /N N 1 =number of “1” bit pixels in a block b T 1,min ; T 1,max upper and lower thresholds N=total number of pixels in the block.

Claim 16

A procedure according to claim 13 , wherein for an 8×8 block the following values hold: T 1,min =55 T 1,max =60 T 2,min =40 T 2,max 32 45.

Claim 17

A procedure according to claim 13 , wherein in blocks with fairly even distribution of bit values in which bit embedding is performed, their neighboring positions have a significant portion of bits of equal value.

Claim 18

A procedure according to claim 13 , wherein in blocks with uniform bit values said embedding takes place in positions in which areas of different values meet so that the bit with the most neighbors of inverse values is changed.

Claim 19

A procedure according to claim 13 , wherein in the case of too large modifications being necessary for said embedding of information in the selected positions, the corresponding block is marked invalid with the invalid areas lying outside the ranges T 1 =(T 1,min ;T 1,max ), T 2 =(T 2,min ;T 2,max ).

Claim 20

A procedure according to claim 17 , wherein an additional tolerance buffer is introduced between T 1 , T 2 and invalid regions.

Claim 21

A procedure for reading marked binary coded data sets, in particular image data and audio signals wherein said binary coded data sets have been marked by embedding at least one information unit in a data set, said procedure comprising the steps of: generating a discrete position sequence using properties characteristic of said data set and using a key; selecting, on the basis of said discrete position sequence, a block of image pixels or audio time signals; transforming said selected block from a spatial or time domain into a frequency domain having a plurality of frequency coefficients; determining the relationship of ones of said frequency coefficients in a predefined frequency range; and recovering said information unit on the basis of said determined relationship.

Claim 22

A procedure for reading marked binary coded data sets, in particular image data and audio signals wherein said binary coded data sets have been marked by embedding at least one information unit in a data set, said procedure comprising the steps of: generating a discrete position sequence using properties characteristic of said data set and using a key; selecting, on the basis of said discrete position sequence, a block of image pixels or audio time signals; transforming said selected block from a spatial or time domain into a frequency domain having a plurality of frequency coefficients; determining the relationship of ones of said frequency coefficients in a predefined frequency range; and recovering said information unit on the basis of said determined relationship.

Claim 23

A procedure for marking binary coded bi-level image data sets by embedding at least one information unit in a data set, comprising the steps of: generating a discrete position sequence using property characteristics of the data set and using a key; selecting, on the basis of said discrete position sequence, a plurality of blocks each containing a plurality of picture elements, each of said picture elements having a value of “1” or “0”; specifying a predetermined relationship between the number of picture elements having a “1” value and the number of picture elements having the “0” value; determining the actual relationship between the number of “1” values and the number of “0” values in said selected block; comparing said determined relationship with said predefined relationship and adjusting the relationship between the numbers of “1” and the numbers of “0” in said block until said determined relationship is equal to said predefined relationship wherein the embedding of an information unit equivalent to a “0” bit encoded in a block b satisfies the following requirement: T 2,min <R(b)<T 2,max with R(b)=N 1 /N N=number of “1” bit pixels in a block N=total number of pixels in the block. T 2,min ; T 2,max upper and lower thresholds.

Claim 24

A procedure for marking binary coded bi-level image data sets by embedding at least one information unit in a data set, comprising the steps of: generating a discrete position sequence using property characteristics of the data set and using a key; selecting, on the basis of said discrete position sequence, a plurality of blocks each containing a plurality of picture elements, each of said picture elements having a value of “1” or “0”; specifying a predetermined relationship between the number of picture elements having a “1” value and the number of picture elements having the “0” value; determining the actual relationship between the number of “1” values and the number of “0” values in said selected block; comparing said determined relationship with said predefined relationship and adjusting the relationship between the numbers of “1” and the numbers of “0” in said block until said determined relationship is equal to said predefined relationship wherein the embedding of an information unit equivalent to a “0” bit encoded in a block b satisfies the following requirement: T 2,min <R(b)<T 2,max with R(b)=N 1 /N N=number of “1” bit pixels in a block N=total number of pixels in the block. T 2,min ; T 2,max upper and lower thresholds.