RU2749880C1 - Method for embedding confidential information in color image - Google Patents

Abstract

FIELD: information security.

SUBSTANCE: invention relates to the field of information security. The selected container is divided into square areas of pixels of a certain size, and the size of the square area of pixels is selected based on the hiding detail. Each subsequent square area of pixels is obtained by shifting the previous square area of pixels to the right by one pixel, and when the border of the image width is reached, the next square area is formed from the beginning of the next line. Next, new square zones are formed by shifting the previous square zone by one pixel until the border of the image width is reached, after which the actions are repeated, while the displacements are performed until the pixels in the selected container are completely enumerated. Then the value of pixels in the square area of pixels is calculated, in each square area of pixels a linear transformation of the elements is performed. The modulus of the determinant of each square area of pixels is determined. The obtained values of the moduli of the determinants of the square areas of pixels are ranked. The information is embedded in the bits of only one pixel of all three color components in a square area of pixels.

EFFECT: increasing the volume of embedded information and resistance to steganalysis methods.

1 cl

Description

The invention relates to the field of information security and can be used for storing and secret transmission of confidential information over open communication channels.

There is a known method of embedding a message in a digital image of the JPEG 2000 format [1], in which a color image for each color component is divided into eight bit layers, and for embedding information, bits are selected at the boundaries of a series of identical bits to minimize distortions and violations of the statistical characteristics of the container.

In this method, there is a need to correct the statistical characteristics, which leads to an increase in the image processing time, as well as to a decrease in resistance to steganalysis and the amount of embedded information.

The closest is the method of embedding a message in a digital image [2], in which the distribution function of the color samples of the pixels of the container is built, then the least significant bit of the number of samples of this container is considered, then the division into blocks is performed, in which the least significant bits of neighboring samples are corrected to write the secret bit of information so that when one of the samples changes by one, the adjacent sample also changes.

The disadvantage of this method is the small amount of information embedding and weak resistance to steganalysis methods.

The technical solution is aimed at increasing the volume of embedded information and increasing resistance to steganalysis methods.

The technical result is achieved by the fact that in the method of embedding confidential information in a color image, which consists in dividing the selected container into zones and embedding information in the least significant bits, the selected container is divided into square zones of pixels of a certain size, and the size of the square zone of pixels is selected based on the detail of the concealment, each subsequent square zone of pixels is obtained by displacing the previous square zone of pixels by one pixel to the right, and when the border of the image width is reached, the next zone is formed by shifting the previous square zone of pixels down by one pixel, while the displacements are made until the pixels in the selected container are completely enumerated, then the value is selected pixels in a square area of pixels, in each square area of pixels a linear transformation of elements is performed, then the modulus of the determinant of each square area of pixels is determined, the obtained values of the moduli of the square determinants are ranked zones of pixels, information is embedded in the bits of only one pixel of all three color components in a square zone of pixels, which is included in the group with the highest value of the modulus of the determinant.

The work of the method can be represented as follows. The selected color image, container, is represented by pixels. All pixels are made up of three color components: red component, green component, blue component (RGB). The entire container is divided into square zones of pixels of a selected or characteristic size. For a more detailed selection of a pixel and a more reliable hiding of information, these square areas of pixels must be selected of the minimum size. For example, two pixels by two. In this case, the number of square areas of pixels in the container is calculated using the following formula:

where h, w are the width and height of the image in pixels.

Calculate the pixel value in each square area of pixels, for example, according to [3]

a _{i, j} = 0.299R + 0.587G + 0.114B,

where i∈0… 1, j∈0… 1 - is used for the case when square blocks of 2 × 2 pixels are used;

R∈0… 255, G∈0… 255, B∈0… 255 are the color components of a pixel, since there are three color components in each pixel.

Now, having the numerical values of each pixel in the square areas of pixels, it is necessary to apply a linear transformation of the values of pixels (elements of determinants) inside each square area of pixels, and this transformation is the same in all square areas of pixels. Then the modulus of the determinant of each square area of pixels is calculated and the obtained values are ranked in descending order. The length of the row varies depending on the choice of the sizes of the square areas of pixels, in our case two by two pixels. Confidential information offered for concealment is embedded in one pixel of each square area of pixels included in the group with the highest value of the modulus of the determinant. Moreover, all three color components of the selected pixel are used.

Thus, the maximum amount of embedded confidential information in the selected container, i.e. color image is calculated as follows:

V = 3 (h-1) (w-1)

As a result, embedded information is almost invisible to most widely used steganalysis methods.

The technical solution provides an increase in the volume of embedded information and resistance to steganalysis methods.

Literature

1. RF patent №2288544, H04L 9/00. A way to embed a message into a digital image of the JPEG 2000 format. 2014.

2. RF patent No. 2262805, H04L 9/00. A way of embedding a message in a digital image ”, 2010.

3. Gonzalez R., Woods R. Digital image processing. - M .: Technosphere, 2005 .-- 1072 p.

Claims (1)

A method for embedding confidential information in a color image, which consists in dividing the selected container into zones and embedding information in the least significant bits, characterized in that the selected container is divided into square zones of pixels of a certain size, and the size of the square zone of pixels is selected based on the detail of hiding, each subsequent square the pixel zone is obtained by shifting the previous square pixel zone to the right by one pixel, and when the border of the image width is reached, the next square zone is formed from the beginning of the next line, then new square zones are formed by shifting the previous square zone by one pixel until it is reached the border of the image width, after which the actions are repeated, while the offsets are made until the pixels in the selected container are completely enumerated, then the value of the pixels in the square zone of pixels is calculated, in each square zone of pixels a linear transformation is performed the development of elements, then the modulus of the determinant of each square zone of pixels is determined, the obtained values of the moduli of the determinants of the square zones of pixels are ranked, the information is embedded in the bits of only one pixel of all three color components in the square zone of pixels, which is included in the group with the largest value of the modulus of the determinant.