MXPA05005390A - Digital image sieves - Google Patents

Abstract

The present invention refers to image sieves based on optical devices containing graphically represented steganographic information, as well as processes for ensuring data safety by this method. The main purpose of the present invention is to provide a data safety device for protecting data elements on physically suitable means useful for storing or displaying graphical information.

Description

"CRIBAS DE IMAGENES DIGITALES" AUTHORS Mat Carlos Galindo Hernández Dr. Guillermo Benito Morales Luna Lie Eduardo Virueña Silva BACKGROUND Currently there are some devices to ensure the authenticity of a printed document and to prevent it from being altered or copied in an unauthorized manner. For this purpose, strategies that involve the introduction of a security code in the document and that act in the arrangement of the pixels that comprise it are usually used. The authenticity of the document is validated through optical resources that amplify these effects and make the security code perceptible. Such devices base their effectiveness on the difficulty that a common user can reproduce the security codes that protect the document without the help of the technology involved. As it is to be supposed, once an unauthorized user has access to such technology through a third party, or because he has been able to infer the strategy of generation of the security codes, the authentication means becomes obsolete and incapable of guaranteeing the integrity of protected documents.

In order to correct these deficiencies, as well as other possible derivatives, a new technological option is presented that uses steganographic characteristics for its implementation and it is precisely this development that is intended to be protected through the present application. We distinguish here the notion of cryptography, as concealment of information, of steganography, as a cover-up of information. This new means of information security also includes other fields of application inherent to its main characteristics, namely: it is based on safety steganographic keys and is independent of a physical effect (such as the refraction of light) for the purpose of validating the authenticity of a document.

DESCRIPTION The three main processes that constitute this technology are described: the creation of the key (which we will call the pixel screen), the codification of the information to be protected and the verification of the authenticity of a generated code.

Consider a secret key composed of an array of N bits (binary digits), organized in m packets of n digits in each packet. A "pixel screen" is then constructed consisting of an image with the key information, arranged as follows: for each of the pixels of the image, which is of size mxn pixels, if the corresponding binary digit in the key is 1, then it turns black and otherwise the pixel does not light up (it remains white). For example, for the key composed of the sequence. { 10011110, 11101101, 01101000, 10110100.}. The resulting pixel screen is the one shown in figure 1. This image printed on a transparent medium constitutes the pixel screen.

The information to be hidden can acquire a great variety of forms, and be of alphanumeric or pictorial type or of almost any type capable of being represented by a distribution of pixels on a digital image. An abstract but illustrative example of a piece of information represented by Figure 2 is presented, which specifically consists of determining which side of the figure is the one that is illuminated (the right one in the case shown).

A new image will be generated, with the information encoded, so that it is unintelligible without the use of the corresponding screen. This new image will be coded under the following criteria: For each pixel (i, j) of the new image, where 1 = i = m, 1 < j < n, one and only one of the following cases is met: a) The pixel (i, j) of the image to be hidden is not illuminated and the pixel (i, j) of the screen is not illuminated either. b) The pixel (i, j) of the image to be hidden is illuminated and the pixel (i, j) of the screen is not. c) The pixel (i, j) of the image to be hidden is not illuminated and the pixel (i, j) of the screen is. d) The pixel (i, j) of the image to be hidden is illuminated and the pixel (i, j) of the screen is also illuminated.

For each of the described cases, we proceed with the pixel (i, j) of the new image according to the following criteria: For the case a) the pixel is illuminated with a very small probability, so as not to substantially add information that can distort the content of the information, being able to take even the value of 0 for such probability, in which case the pixel definitely does not light up. For case b) the pixel is illuminated with a very high probability, so as not to substantially eliminate the content of the information, and such probability can even take the value 1, in which case the pixel is definitely illuminated. For the case c) the pixel is illuminated with a balanced probability with the sole purpose of adding noise on the information of the new image. Considering that all these added pixels will be hidden by the screen, the probability of illuminating the pixel can be weighed with a certain proportion, thereby causing the noise to be better distributed instead of concentrating on the information areas, as will be appreciated later in Some examples. For the case d) the pixel is illuminated with a balanced probability as in the previous case, because such pixels will also be hidden by the superposition of the screen. Optionally, the probability of not illuminating the pixel can be increased to obtain precisely the same effect as the previous point. This is how the distribution of illuminated pixels in the areas where the original information is concentrated is being regulated.

To illustrate this process, see Figure 3 which shows the screen (colored in light tone) superimposed on the original information to be hidden. Figure 4 shows processes a) and b) applied in the strict sense, that is with probabilities 0 and 1 respectively. The above is only motivated by the restrictions (size and amount of information) in this illustration. Figure 5 shows the result of applying step c) with a high probability of illuminating the segments that do not contain original information (colored in dark tone to facilitate their perception). Figure 6 shows the effect of applying criterion d) with a minimum probability of illuminating the segments that already contained information.

The resulting image (illustrated in Figure 7) is referred to as the security code and completely hides the original information. From the security code, that is, without the knowledge of the screen, it is impossible to determine which of the two sides was originally illuminated.

To be able to read the original information (know which side was originally lit), it is sufficient to superimpose the corresponding pixel screen in an exact way on the coded image, as illustrated in figure 8 (the screen is illuminated in a light tone for facilitate their identification). After observing the screen superimposed on the encoded information, it is immediate to know which side of the original figure was the illuminated one.

Once the abstract procedure of the three main processes has been described, namely the generation of the screen, the coding of the information to be protected and the validation, some more descriptive results about the application of the technology are illustrated.

Figures 9 through 14 show examples where different probability values are used to execute steps a), b), c) and d) defined above. For each illustration, the coded key (the pixel screen), the information to be hidden, the result of applying step a), the result of applying step b), the result of applying the step c), the result of applying step d), and the image obtained from superimposing the screen on the obtained code (validation process) The probabilities that were used for the execution of steps a), b), c) and d) in each illustration were: Figure 9: a) 0.0, b) 1.0, c) 0.5 d) 0.5 Figurra a) 0.3, b) 1.0, c) 0.7 d) 0.3 Figure 11 a) 0.1, b) 0.9, c) 0.9 d) 0.1 Figure 12 a) 0.3, b) 1.0, c) 0.9: d) 0.2 Figure 13 a) 0.1, b) 1.0, c) 1.0 d) 0.1 Figure 14: a) 0.3, b) 1.0, c) 0.8 d) 0.1 All the examples were made with screens of 48 x 48 pixels dimension and with the same key for the screen, as well as the same information to be codified to facilitate comparison.

Figures 15 to 20 show examples where the variation of the security codes generated under the same pixel screen is illustrated, when applied to different information to be hidden. The probabilities used for processes a), b), c) -d) were 0.1, 1.0, 1.0 and 0.1 respectively, with the sole purpose of illustrating the readability of the information.

Figures 21 to 23 show examples showing the variation of the security codes generated for screens of different sizes. The probabilities used for the execution of processes a), b), c) and d) were similar to those of the previous example.

The Screens of Images find use in different fields of application. One of the simplest examples is to validate a printed document with a security code (figure 24) by overlaying the corresponding pixel screen (figure 25) to the document in question. In this case, the sieve may well simply consist of an acetate card printed with the coded key.

Patently, sensitive information contained in the normal document should be hidden in the security code, which should allow detecting any alteration to the information of the same. Another use detected for this technology is that it serves as an additional security means to validate a user's identity on an Internet site. When a user, previously subscribed to a portal (financial or bank, for example) is identified on the site, the system shows a security code such as those already described (figure 27, item 1). This code can only be deciphered by the pixel screen, which in the process of enrollment to the site, has been delivered to the user (figure 27, item 3). It should be mentioned that this security scheme eliminates the risk of fraud in banking operations carried out in Internet cafes that are operated by "hackers" capable of seeing what users type and the information they receive. In this scheme, the "hacker" in question can obtain the image of the code and the operation passwords to carry out financial transactions on the site, however, he will not be able to make use of them when he is unable to decipher the security code, only visible by superimposing the corresponding screen, as shown in figure 27 (item 2).

Three more examples of the use of the screens are attached in figures 28, 29 and 30, visible only with the use of the attached screen printed in figure 31 on acetate for validation.

The use of this technology has applications in various fields, thanks to the ease of recognizing the printed codes in virtually any surface, not necessarily flat, for example, in spherical, cylindrical or other similarly diverse shapes and materials.

Several aspects of this technology have been observed, based mainly on the parameters described below, classified mainly by their object of application: Space of functions: The criteria for illumination or coloring of the pixels are not restricted to the four basic principles mentioned in the illustrations, but to the entire space of functions that have as domain the information represented by the image to be hidden and as a counter-domain the set of all the possible security codes to generate, not necessarily of the same size or quality of information as those of its counterpart. Consider, by nominating a simple example, a function based on the distance (Euclidean) closest to each pixel to the source of information to be hidden, as a criterion for the illumination of the corresponding pixel in the security code, which is in itself a case generalized of the four basic principles that were taken as illustration.

Superposition: 2 or more screens of images are incorporated at the same time to be able to decipher a given code. In the same way two or more encodings of the information to be hidden can be superimposed to provide obfuscation effects or some other similar one.

Biased: statistically biased keys are generated, that is, they have a greater or lesser number of 1's than of 0's. Consider also the characteristic of skewing from the point of view of the graphic concentration of illuminated pixels, allowing the generation of clusters arranged by some particular criterion.

Coloring: colored pixels are used on the screens, either to have the effect of simple "noise" on the coding of the screen or the code, or to make additions of frequency (blue + yellow = green, yellow + red = orange , etc.) that give a different nuance to the validations of the codes.

Refraction: Screens are printed on lenses or on objects capable of causing refractive effects of light, using codes previously prepared for that purpose.

Projection: The screens are composed of light projection devices on surfaces that contain the security code in question.

Stereograms and anaglyphs: The validation means incorporate stereograms or anaglyphs as devices to encode the information to be hidden.

Technological Adaptation: It is observed the integration of technological devices already known that by their nature allow or contribute to the decoding of security codes. Take, for example, the adaptation of a laser reading device (similar to those used for reading barcodes) adapted to read the arrangement of the pixels that make up a security code, or an electronic diary capable of reading codes. security from a computer and display the deciphered codes or assemblies required for decoding on a screen. Note not only the integration of physical devices, but also logical means, algorithms or processes of a non-tangible nature that can be incorporated into the use of this technology. Take as an example for this case, the use of pattern recognition algorithms or artificial intelligence to treat the reading of decoded security codes.

All these technological benefits allow this means of security to far exceed the scope achieved by other devices that pursue the same purpose, resulting steganographically and cryptographically more robust and with a greater application scope.

Claims (3)

CLAIMS Having sufficiently described my invention, I consider it a novelty and therefore claim as my exclusive property what is contained in the following clauses:

1. Image screens consisting of graphic information sources associated with key steganographic nature, stored on one or more optical devices that visually operate with specially encoded images, allow the readability of hidden information.

2. Coding algorithms for image screens and security codes, based on pseudo-random and random disposition criteria of the information composed of the data to be hidden together with the steganographic key (s) stored in the screens in question, as well as the variations inherent to the application of physical or logical properties such as the superposition or combination of screens, the generation of graphic and statistically biased screens, the use of colors in security codes or in image screens, or in both, the integration of elements with optical effects such as refraction or projection of light beams, or for the codification of stereograms or anaglyphs or the use of any function for the distribution of pixels over the security codes, contemplated in the space that has as a domain, the information to be represented represented graphically and digitally and as a contradictory the conjunct of possible security codes represented in the same way, but not necessarily of the same graphic nature. Of course, consider our exclusive property the use of any of the combinations of elements already defined.

3. Adaptation of technological elements already known, of a physical or logical nature for any of the processes directly or indirectly associated with the use of this technology, namely generation of security keys, representation of the information to be hidden, generation of security codes and decoding of hidden information.