WO2004100527A1

WO2004100527A1 - Image processing device, image processing method, image processing program, and mechanically readable recordoing medium containing the image processing program

Info

Publication number: WO2004100527A1
Application number: PCT/JP2004/005586
Authority: WO
Inventors: Kazuyuki Nako
Original assignee: Sharp Kabushiki Kaisha
Priority date: 2003-05-09
Filing date: 2004-04-19
Publication date: 2004-11-18
Also published as: JP2004336539A

Abstract

A printer for processing an image to be output generates a random dot image where points are arranged at random (S10-S12) and performs exclusive OR calculation between each corresponding pixels between the generated random dot image and a first mask image (32) prepared in advance, thereby generating an encrypted image (S13). The generated random dot image and the encrypted image are subjected to processing such as geometric conversion so that the corresponding pixels are matched when they are superimposed and printed out (S15). When the images are superimposed by the exclusive OR calculation and output, the first mask image (32) embedded in the encrypted image via the random dot image appears visually. Only the image in which the first mask image (32) can be viewed is authenticated to be a genuine one. Since this superimposing accuracy cannot be obtained by a normal copying machine, it is possible to prevent forgery.

Description

TECHNICAL FIELD The present invention relates to an image processing apparatus, an image processing method, an image processing program, and a machine-readable recording medium storing the image processing program.

The present invention relates to an image processing apparatus, an image processing method, an image processing program, and a machine-readable recording medium storing the image processing program, and more particularly to an image processing apparatus, an image processing method, and an image processing method for embedding another image in an output image. The present invention relates to a machine-readable recording medium that stores a processing program and an image processing program. Background art

Recent advances in copier technology have made it possible to easily copy and forge documents, which has become a major social problem. For this reason, image embedding technology is one of the technologies that makes it easy to distinguish a genuine product from a duplicate and makes it difficult to forge.

Traditionally, banknotes are embedded with pictures such as portraits and buildings as watermarks to prevent counterfeiting.However, special paper called woven paper is required. It could not be applied to paper and copiers. In the method of generating a show-through pattern and the printing apparatus, general paper is arranged by arranging the first halftone pattern and the second halftone pattern so as to form markings that can be visually recognized by illumination from the show-through light source. And a technology for realizing a watermark with a copying machine has been provided (for example, see Japanese Patent Application Laid-Open No. 2002-142105). In this technology, “watermarking” is realized by locally shifting at least one of the phase, angle, and frequency of the halftone dot of the second halftone pattern. As a result, the “watermark” is reproduced even if accurate alignment is not performed on both sides, but it can be easily duplicated, and the original effect of the watermark and the effect of preventing forgery were poor. Also, since the second halftone pattern is locally deformed, there is also a problem that it is easy to visually recognize what watermark is embedded simply by observing the second halftone pattern. Was. It is not possible to see exactly what is embedded from only one image, and as a technique to make the embedded image appear only by superimposing the two images, copyright such as the signature of the image author There is a method for embedding information related to the above in an image and a copying apparatus for embedding copyright information in an image (for example, refer to Japanese Patent Application Laid-Open No. Hei 9-252,397). In this copying apparatus, a density pattern method of expressing density by n * n pixels is applied, and the redundancy of a combination of pixels when expressing the same density is used. In other words, two patterns are prepared according to the luminance level. Then, different patterns are assigned to pixels where information is embedded, and the same pattern is assigned to pixels where information is not embedded. Therefore, the image to be embedded is in units of η * η pixels. However, in order to verify the embedded image, it is necessary to create a key verification image created at the time of embedding, and the problem of storing and delivering the verification image remains. There was a problem with convenience. In addition, since the density pattern requires a size of at least 2 × 2 pixels, there is a problem that the resolution of the image is lower than when information is embedded in pixel units.

Some of the technologies that can be quickly identified as a copy when copied are as follows. In other words, in a printed matter with a latent image suitable for preventing forgery by copying, the latent image is printed with fine halftone dots, and the white background is printed with coarse halftone dots that look uniform, so that when the printed matter is copied, the latent image is printed. There has been proposed a technique that can be clearly recognized (for example, see Japanese Patent Application Laid-Open No. 54-74125). This uses the limit of the reproduction performance of a copier, and the same technology is applied to copying of a resident's card, etc., so that the word “copy” emerges when copied.

However, it was difficult to prevent forgery because the same device and method could be used to make the same product, and it was difficult to distinguish it from the real product. In addition, it was difficult to recognize at first glance that the image was real because the image emerged only after copying. Therefore, it has been desired to provide a function that is difficult to forge and can easily verify the authenticity, that it is difficult to make a duplicate, and that it is easy to determine that the duplicate is a duplicate even if it is duplicated. Disclosure of the invention An object of the present invention is to provide an image processing apparatus, an image processing method, an image processing program, and a machine-readable recording medium storing the image processing program, which can easily verify that the image is a real image.

Another object of the present invention is to provide an image processing apparatus, an image processing method, an image processing program, and a machine-readable recording medium that stores the image processing program, which makes it difficult to forge an image. .

In order to achieve the above object, according to one aspect of the present invention, an image processing apparatus for processing an image to be output includes a first random dot for generating a first random dot image in which points are randomly arranged. An image generating means for generating a second random dot image by performing a predetermined operation for each pixel corresponding to the generated first random dot image and a previously prepared first mask image; And two random dot image generating means. Then, the predetermined operation is performed when the generated first random dot image and the second random dot image are superimposed on each other so that the pixels are coincident with each other and output, via the first random dot image. This is an operation to make the mask image appear visually.

Therefore, only the person or device that has (can generate) the first random dot image used when generating the second random dot image has the second random dot image received from the other party ( The first mask image can be restored by performing the reverse operation of the predetermined operation on the generated (first generated) random dot image. As a result, encrypted communication using the first mask image as secret information becomes possible. In addition, the first random dot image that has (generated) is printed on a transparent sheet, and the second random dot image received from the other party is printed on the printed first random dot image via the transparent sheet. When the images are superimposed, the inverse operation described above has been performed, and the first mask image can be restored. As a result, encrypted communication using the first mask image as secret information becomes possible.

In this way, when the second random dot image is superimposed on the first random dot image, only the first random dot image that allows the first mask image to appear visually is used as the real ( Not forged). Also, do the superposition of both require that the corresponding pixels match? Therefore, duplication of the output result can be made difficult, and forgery can be prevented.

In addition, the second random dot image for visually appearing the first mask image via the first random dot image is generated using the first random dot image generated each time. Therefore, forgery can be more reliably prevented. Also, since the second random dot image is generated using the first random dot image, there is no need to store the second random dot image.

Preferably, a superimposition processing means for processing the generated first random dot image and the second random dot image so that the corresponding pixels coincide when both are superimposed and output. Further prepare.

Preferably, the above-mentioned predetermined operation is an exclusive-OR operation, so that the first mask image can be embedded in the second random dot image by a simple operation for visual appearance. In this case, the reverse operation described above is a logical sum operation.

In addition, the second random dot image is read by a scanner or the like, and the first mask image is completely restored by performing an exclusive OR operation on the read second random dot image and the first random dot image. You can also.

Preferably, the first random dot image generating means generates the first random dot image such that the appearance probability of a point is approximately 50%.

Therefore, even in the generated second random dot image, the appearance probability of points is approximately 5

Since it is 0%, it just looks like points are randomly arranged. As a result, the first mask image embedded in the second random dot image can be hidden from a third party.

Preferably, the first random dot image generating means generates the first random dot image according to a random number sequence uniquely determined using the key data, so that the first random dot image can be easily generated. Hiding the data from a third party can prevent forgery of the first random dot image. In addition, since the first random dot image can be generated by using the key data, it is not necessary to store the first random dot image. Also, when the second random dot image is transferred, the first mask image can be restored by transferring the key data.

Third parties who do not know the key code must decode the _second random dot image. You can't, and you can't fake the same. Since only the person who knows the key data can restore the first mask image, it is easy to verify that the second random dot image is genuine, and it is difficult to forge the second random dot image. Preferably, the key data includes a different identifier generated when outputting the image. If the same key data is used continuously, the key data does not need to be changed, so the convenience is high. Generally, the encryption strength (the degree of difficulty in decrypting the encryption) decreases. Therefore, by including an identifier in the key data, a different first random dot image is generated if the identifier is different, so that the encryption strength is not impaired.

Preferably, the image processing device further includes data input means for externally inputting the key data. Therefore, the key data can be arbitrarily changed from the outside, and a decrease in encryption strength can be suppressed.

Preferably, the identifier is output together with the second random dot image. Therefore, even if a second random dot image is generated using a different first random dot image for each image output, the identifier included in the key data is output together with the second random dot image. Therefore, random dots can be reproduced by using the output identifier, so there is no problem in decoding the second random dot image.

Preferably, an image embedding unit is further provided. The image embedding means sets each point of at least one of the first random dot image and the second random dot image at a different density and the same density according to the second mask image prepared in advance. The second mask image is embedded in at least one of the first and second random dot images by converting the image into a pattern that is easily visible.

Therefore, two images of the first mask image and the second mask image can be embedded in the random dot image. The first mask image appears when the first and ^second random dot images are superimposed, and for each corresponding point in the random dot image in which the second mask image is embedded, the second mask image is used. Is converted to a pattern that looks different in density and has the same density. As a result, the second mask image can appear by copying the random dot image with a copying machine or the like. Therefore, the appearance of the second mask image identifies that the image is a fake copied by a copying machine or the like and not a genuine one (not copied). it can.

Preferably, when the image is printed out on paper, the first random dot image is printed out on one side of the paper and the second random dot image is printed out on the other side of the same paper.

Therefore, since the first random dot image is printed on one side of the printing surface of the paper and the second random dot image is printed on the other side, the printed material (paper) is held over a light source to allow light to pass therethrough. Can appear.

Preferably, the superposition processing means has a mirror image conversion means. Since the mirror image conversion means converts one of the first random dot image and the second random dot image into a mirror image with respect to the other, the overlay can be performed accurately. Preferably, the overlay processing means has an image conversion means. The image conversion means performs at least one of a translation conversion, a rotation conversion, and a scaling conversion on at least one of the first random dot image and the second random dot image. Therefore, the image is output after detecting the positional deviation during double-sided printing on paper in advance, and the image can be printed at a more accurate overlapping position. Therefore, it is edible g to make the first mask image (watermark) clearly appear in the printed image.

In general, copiers that can print on both sides do not completely match the print positions on the front and back due to the paper feeding accuracy of the paper and the mounting accuracy of the print drum. It is difficult to reproduce the same “watermark” even if the three parties copy on both sides. Therefore, whether or not the image is a forged image can be quickly determined based on whether or not the first mask image (watermark) appears in the image of the copy result, and the forgery action can be suppressed.

Preferably, the image processing device further includes an image density conversion unit. The image density conversion means changes the density of at least one of the random dot images so that the image density of the first random dot image and the second random dot image are different depending on the light transmittance of the printing surface of the paper. Perform the conversion process.

Therefore, the density of the random dot image on the side where the 'watermark' (first mask image) is observed is made lighter than that of the random dot image with the 'watermark' embedded. By doing so, it is possible to make the density of the random dot image on the observing side equal to the density of the random dot image on the back side (the opposite side to the observing side) that can be seen transparently, making it possible to observe the watermark more clearly. Become.

To achieve the above object, according to another aspect of the present invention, there is provided an image processing method for processing an image to be output, comprising: a first random dot image for generating a first random dot image in which points are randomly arranged; An image generation step, and a second operation of generating a second random dot image by performing a predetermined operation for each pixel corresponding to the generated first random dot image and the previously prepared first mask image And generating a random dot image. Then, the predetermined calculation is performed when the generated first random dot image and the second random dot image are superimposed on each other so that the pixels coincide with each other and are output, via the first random dot image. This is an operation for visually appearing one mask image.

Therefore, only the person or device that has (can generate) the first random dot image used to generate the second random dot image has the second random dot image received from the other party ( The first mask image can be restored by performing the reverse operation of the predetermined operation on the generated (first generated) random dot image. As a result, encrypted communication using the first mask image as secret information becomes possible. In addition, the first random dot image that has (generated) is printed on a transparent sheet, and the second random dot image received from the other party is printed on the printed first random dot image via the transparent sheet. If the images are superimposed, the above-described inverse operation has been performed, and the first mask image can be restored, so that encrypted communication using the first mask image as secret information becomes possible.

In this way, when the second random dot image is superimposed on the first random dot image, only the first random dot image that allows the first mask image to appear visually is used as the real ( Not forged). Also, since the superposition of the two requires that the corresponding pixels match, it is possible to make it difficult to duplicate the output product and prevent forgery.

In addition, the second random dot image for visually appearing the first mask image via the first random dot image is a first random dot image generated each time. Since the image is generated using the cut image, forgery can be more reliably prevented. Also, since the second random dot image is generated using the first random dot image, there is no need to store the second random dot image.

Preferably, the method further comprises a superposition processing step of processing the generated first random dot image and the second random dot image so that when the two are superimposed and output, the corresponding pixels coincide. .

Preferably, the predetermined operation is an exclusive OR operation.

Preferably, in the first random dot image generating step, the first random dot image is generated such that the appearance probability of the points is approximately 50%.

Preferably, the first random dot image generating step generates a first random dot image according to a random number sequence uniquely determined using the key data.

Preferably, the key data includes a different identifier generated each time an image is output.

Preferably, the image processing apparatus further includes a data input step of inputting the key data from outside.

Preferably, the identifier is output together with the second random dot image.

Preferably, the method further includes an image embedding step. In the image embedding step, each point of at least one of the first random dot image and the second random dot image is set at the same density at a different density according to the second mask image prepared in advance. A process of embedding the second mask image in at least one of the first and second random dot images is performed by converting the pattern into a pattern that looks like a density. Preferably, when the image is printed out on paper, the first random dot image is printed out on one side of the paper and the second random dot image is printed out on the other side of the same paper.

Preferably, the superposition processing step includes a mirror image conversion step. In the mirror image conversion step, one of the first random dot image and the second random dot image is converted into a mirror image with respect to the other.

Preferably, the superposition processing step includes an image conversion step. The image conversion step consists of reducing the number of the first random dot image and the second random dot image. Apply at least one of translation, rotation and scaling to at least one image.

Preferably, the image processing method further includes an image density conversion step. In the image density conversion step, the density of at least one of the random dot images is changed so that the image density of the first random dot image and the second random dot image differs according to the light transmittance of the printing surface of the paper. Perform the conversion process.

An image processing program according to still another aspect of the present invention is a program for causing a computer to execute the above-described image processing method.

A machine-readable recording medium according to still another aspect of the present invention is a recording medium recording the above-described image processing program. BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 is an external view of a printer that realizes an image output device according to an embodiment of the present invention.

2A and 2B are diagrams showing an example of the control block and memory contents of the printer shown in FIG.

FIG. 3 is a configuration diagram of a computer according to the embodiment of the present invention.

FIG. 4 is a flowchart illustrating a control structure of a program executed by the image output device according to the embodiment of the present invention.

FIG. 5 is a diagram illustrating an example of a random dot image.

FIG. 6 is a diagram illustrating an example of a first mask image.

FIG. 7 is a diagram illustrating an example of an encrypted image.

FIG. 8 is a diagram illustrating an example in which an encrypted image and a random dot image are superimposed. FIG. 9 is a diagram illustrating the principle of duplex printing in a printer.

FIG. 10A and FIG. 10B are diagrams illustrating density conversion.

FIG. 11 is a diagram illustrating a method of embedding the second mask image.

FIG. 12A and FIG. 12B are diagrams showing application examples according to the present embodiment.

FIGS. 13A and 13B are diagrams showing application examples according to the present embodiment. BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the following description, the same parts are denoted by the same reference numerals. The names and functions are the same. Therefore, detailed description thereof will not be repeated.

Processing in the image output apparatus according to the present embodiment is realized by software executed by an image processing control engine such as a printer or a copier having both a printer function and a scanner function. It may be realized by printer driver software executed on a personal computer or a workstation, or may be realized as dedicated hardware. Further, the printer may be a device including a scanner unit like a general copying machine.

FIG. 1 shows the appearance of a printer 1 as an example of an image output device. Referring to FIG. 1, the printer 1 is realized by an operation panel 13 having an operation unit for inputting information from outside and a display unit for displaying information to the outside, an Ethernet (R), etc. Includes external I / F (Inter Face) 14 for input / output or communication with external devices that do not have any.

FIG. 2A shows a block configuration of the printer 1. The printer 1 has a CPU (Central Processing Unit) 10 and a program for centrally controlling and managing the printer 1, which is connected to the operation panel 13 and external I / F 14 via a bus 16 And a ROM (Read Only Memory) 11 for storing data and data, a RAM (Random Access Memory) 12 and a print engine 15 for printing image data.

The CPU 10 includes a first random dot image generation unit 17, a second random dot image generation unit 18, a superimposition processing unit 19 having a mirror image conversion unit 20 and an image conversion unit 21, and an image embedding unit. 22 and an image density converter 23. The functions of these units are realized by programs stored in the RAM I2 or ROM 11 in advance, and the CPU 10 reads and executes each program.

RAM I 2 is not only used as a storage area for programs and data, but also as a work area necessary for executing programs. Referring to FIG. 2B, the data stored in the data storage area of RAM I 2 includes a random dot image 30, a symbol image 31, and first and second mask images 32 and 3 described later. 3, surface And back side print data 34 and 35, variable-size random dot image 3 OH :, variable-size signal image 31 H, geometric random-dot image 30 G, dark-variable random dot image 30 D, dark-variable encrypted image 31 D, synthetic encrypted image 36 and the data of the composite random dot image 37 are included. Since the random dot image 30 refers to an image in which dots are randomly arranged, and the encrypted image 31 is also treated as an image in which dots are randomly arranged, if the random dot image 30 is set as the first random dot image, the encrypted image 3 1 is the second random dot image.

As described above, the image output device according to the present embodiment is realized by the hardware of the printer 1 and the software executed by the CPU 10. The software executed in the printer 1 is stored in the ROM 11 in advance, read out from the ROM 11 by the CPU 10 and loaded into the RAMI 2 to be executed.

The case where the image output processing according to the present embodiment is realized by printer driver software executed on a personal computer or a workstation is as follows. FIG. 3 is a configuration diagram of a computer having an image output processing function according to the present embodiment.

Referring to FIG. 3, the computer includes a monitor 110 such as a cathode ray tube (CRT), a CPU 122 for centrally controlling the computer itself, a memory 124 including ROM or RAM, a fixed The disk 126 and FD (flexible disk) 132 are detachably mounted, and the FD drive 130 and the CD-ROM (Compact Disc Read Only Memory) 142 that access the mounted FD 132 are detachably mounted. And a communication interface 180 for communicatively connecting the computer with the CD-ROM drive 140, the keyboard 150, the mouse 160, the pentagram 170 and the communication network 182 for accessing the mounted CD-ROM 142. These units are communicatively connected via a bus. The computer may be provided with a magnetic tape device (not shown) for detachably mounting a cassette type magnetic tape and accessing the magnetic tape.

The output information (image) as the execution result of the printer driver software of the computer in FIG. 3 is printed by a printer (not shown). This printer connects to a computer over a communication network 182 via a communication interface 180 Or may be connected via a bus.

The printer driver software is stored on a computer-readable recording medium. In the present embodiment, as the recording medium, a memory required for performing the processing in FIG. 3, for example, a ROM itself of the memory 124 may be a program medium, or an external storage device. A program reading device such as a magnetic tape device and a CD-ROM drive device 140 may be provided, and a magnetic tape or a CD-ROM 142 as a recording medium may be inserted into the program reading device to read the program. . In any case, the stored program may be configured to be accessed and executed by the CPU 122, or in any case, the program is read and read, and the read program is 3 may be loaded into a predetermined program storage area of a computer, for example, a program storage area of a RAM of the memory 124, and read and executed by the CPU 122. It is assumed that the loading program is stored in the computer in advance.

Here, the above-described program medium is a recording medium that is configured to be separable from the computer main body, and may be a medium that fixedly holds a program. For example, tapes such as magnetic tapes and cassette tapes, magnetic disks such as FD132 and fixed disk 126, CD-ROM 142 / MO (Magnetic Optical Disc) / MD (Mini Disc) / DVD (Digital Versatile Disc) ) And other optical discs, IC (Integrated Circuit) cards (including memory cards), card systems such as Z optical cards, or mask ROMs, EP ROMs (Erasable and Programmable ROMs), EEPROMs (Electrically EPROMs), This is a semiconductor memory using a flash ROM or the like.

Further, in the present embodiment, since the computer is configured to be connectable to communication network 182 including the Internet via communication interface 180, the program is downloaded from communication network 182. As described above, the medium may carry the program in a fluid manner. When the program is downloaded from the communication network 182 in this manner, the download program is stored in the computer in advance or by another It is assumed that it is installed in the computer main body in advance from a recording medium. Note that the content stored in the recording medium is not limited to a program, but may be data.

The printer 1 and the workstation shown in FIGS. 1 and 2A, and the hardware itself of the computer shown in FIG. 3 are common. Therefore, the most essential part of the present invention is the software recorded on a recording medium such as CD-ROM 142, FD 132, ROM 11 and RAMI 2.

The well-known portions of the operation of the printer 1 itself shown in FIGS. 1 and 2A will not be described in detail.

With reference to the flowchart of FIG. 4, a procedure executed by the image output device according to the present embodiment will be described. First, since the user inputs a key code using the operation panel 13, the CPU 10 receives the input key code in step (hereinafter, step is abbreviated as “S”) 10. It is not necessary for the user to input the key code every time, and the input key code may be stored in the RAMI 2 and the CPU 10 may read the key code stored in the RAMI 2 . When the image output device according to the present embodiment is realized by printer driver software, operation panel 13 can be replaced with a GUI (Graphical User Interface) and a keyboard.

In S11, CPU 10 generates a different identifier each time printer 1 prints out an image. The identifier may be a random value or a serial number value. Here, it is sufficient that at least one of S10 and S11 is executed.

In S12, the first random dot image generation unit 17 of the CPU 10 generates a pseudo random number sequence using the data serving as the “key”, and performs a predetermined procedure using the original random dot image and the generated pseudo random number sequence. , A random dot image 30 in which “points” are randomly arranged is generated, and the generated random dot image 30 is stored in the RAMI 2. Here, the key data is a combination of the identifier and the key code hidden from the third party. Here, the “point” corresponds to one black pixel, but the number of corresponding pixels is not limited to one, and the shape is a rectangle having a certain area. Or a circle. The color is not limited to black, but may be gray or red. By generating a pseudo-random number sequence using a combination of a key code and an identifier as a “key”, a pseudo-random number sequence uniquely determined from the key code and the identifier can be generated.

Since the algorithm for generating the pseudo-random number sequence is well known, its description is omitted. Preferably, by generating the random dot image 30 so that the probability of appearance of points in an image area of a predetermined size of the random dot image is approximately 50%, the following symbol image generated from the random dot image 30 Even 3 1 looks like a random array of points, making it difficult for a third party to know what is embedded. That is, since the encryption image 31 is generated by an exclusive OR operation between the pixels of the random dot image 30 and the first mask image 32, the encryption image 31 is set to approximately 50%. Even when is output, it can be difficult to determine where the first mask image 32 is embedded in the encrypted image 31.

In S13, the second random dot image generation unit 18 of the CPU 10 generates the random dot image 30 stored in the RAM I2 and the first mask image previously stored in the RAM I2. 3 2 is read out, and the value of each pixel of each image is exclusive-ORed to generate a symbol image 3 1 which is a random dot image, and the generated encrypted image 3 1 is converted to a RAM I 2 To be stored. Although the exclusive OR operation is used here, the present invention is not limited to this. In other words, any operation that enables the first mask image to appear visually through the first random dot image when the random dot image and the encrypted image are superimposed and printed out is sufficient. . For example, a type of operation that replaces the image of the partial area corresponding to the first mask image 32 of the random dot image 30 with another random dot image may be used.

In S14, the CPU 10 scales the random dot image 30 and the encrypted image 31 stored in the RAM 2 into a scaled variable, and changes the scaled random dot image 30H and the scaled image. 31 Stored in RAM I 2 as 1H. This is because printing a random dot image 30 or encrypted image 31 at a 1: 1 ratio with the resolution of the printer 1 will destroy points (pixels) and make it very difficult to align the positions of both images. For reasons. The scaling ratio is 5 when the resolution of printer 1 is 600 DPI (Dots Per Inch). About twice is appropriate. If the image is too large due to scaling, it will be easier to align, but it will also be easier to duplicate. Conversely, if it is too small, the duplication will be difficult, but the alignment will also be difficult. Such scaling conversion of an image according to the resolution of the printer 1 is a general image processing, and therefore the description is omitted.

In S15, the superimposition processing unit 19 of the CPU 10 performs one or more geometric transformations on the variable-size random dot image 30H stored in the RAM I2 and performs a geometric random dot image. Generate 30 G and store it in RAM I 2. The one or more geometric transformations performed here include a mirror transformation by the mirror transformation unit 20 and at least one of a translation transformation, a rotation transformation and a scaling transformation by the image transformation unit 21. This geometric transformation is a process for matching (matching) the positions of the corresponding pixels when the variable-size random dot image 30H and the variable-size encrypted image 31H are superimposed.

It should be noted that the geometric transformation may be performed on the zoomed-in symbol image 31H. In other words, at least one of the variable-size random dot image 30H and the variable-size encrypted image 31H may be used.

The geometric transformation may include a linear transformation, which will be described later.In this case, a single transformation should be performed by combining the mirror image transformation, the translation transformation, the rotation transformation, and the scaling transformation, respectively. Can be.

If only the encrypted image is to be printed without performing double-sided printing, S15 may be omitted. Here, printing only the encrypted image means that only the person or device that has (reproduces) the random dot image used to generate the encrypted image can restore the first mask image 32, This makes it possible to perform encrypted communication using the mask image 32 as secret information.

In S16, the image density conversion section 23 of the CP 10 performs density conversion on the geometric random dot image 30G and the variable-magnification encryption image 31H stored in the RAM I2. The dark-density random dot image 30 D after replacement and the dark-density encrypted image 31 D are stored in RAM 12. Density conversion can be performed by converting point (pixel) data into a pseudo-shade pattern (a pattern that has a different pixel density from the target and looks the same density). At this time, according to the light transmittance of the printing surface, for example, In this way, a pseudo shade pattern is selected and the density is converted. As a result, when the encrypted image is observed from the side where the image is printed, the density of the encrypted image is close to the density of the random dot image that can be seen through light, and the “watermark” (first mask image) is more clearly seen. can do.

Also, the image embedding unit 22 of the CPU 10

Reads the second mask image 33 and converts the density of the random dot image and the symbol image so that different pseudo gray patterns are used at points (pixels) corresponding to the second mask image 33 and at points (pixels) not corresponding to the second mask image 33 May be. This allows the second mask image 33 to appear when copied. When outputting only the encrypted image, the density conversion processing of the random dot image may be omitted. Further, when the second mask image 33 is not embedded, the process of applying the pseudo shade pattern according to the second mask image 33 may be omitted.

In S17, the CPU 10 reads out the converted dark-symbol image 31D stored in the RAMI2 and the front side print data 34 and combines them, and stores the resultant encrypted image 36 in the RAMI2. At this time, the identifier generated in S11 may be read from RAMI2 and combined with the front side print data 34. Compositing refers to the process of compositing images. The printer 1 prints together the compositing targets (images, data, identifiers, etc.).

In S18, the CPU 10 reads and combines the dark-density random dot image 3 OD after conversion stored in the RAMI 2 and the back side print data 35, and combines the resultant random dot image 37 thus obtained with the RAMI 2. Store in 2. If the double-sided printing is not performed, the process of S18 can be omitted. In the combination of S17 and S18, a combination of the print data 34 and the encrypted image 31D and a combination of the print data 35 and the random dot image 30D are used, but the combination is not limited to this. That is, the print data 34 and the random dot image 30D may be combined, and the print data 35 and the symbol image 31D may be combined.

In S19, the CPU 10 prints the combined encrypted image 36 and the combined random dot image 37 stored in the RAMI 2 on the front and back sides of the same printing paper using the printing engine 15. Compositing stored in RAMI 2 when not performing duplex printing Print only the encrypted image 36.

The random dot image 30 generated in S12 will be described with reference to FIG. In order to generate the random dot image 30, a pseudo-random number sequence is generated using a combination of a key code and an identifier as a key. For example, the value of the pseudo-random number sequence is made to correspond one-to-one with the pixels that make up the random dot image 30.Each pseudo-random value is divided by 2, and if the remainder is 1, the corresponding pixel is scored. By generating and not generating points when the remainder is 0, it is possible to generate a random dot image 30 with a point appearance probability of about 50% for each pixel as shown in Fig. 5. it can. If the size of the random dot image 30 is fixed, there is a one-to-one correspondence between the pseudo-random number sequence and the random dot image 30. Therefore, if only a key code and an identifier are given at the time of decoding, a unique random dot image 30 is reproduced. (Decryption).

With reference to FIG. 6, FIG. 7 and FIG. 8, the process of generating encrypted image 31 performed by CPU 10 in S13 will be described. For example, in the case of the first mask image 32 shown in Fig. 6,

The encrypted image shown in FIG. 7 is generated by performing an exclusive OR operation on the random dot image 30 shown in FIG. 5 and the first mask image 32 shown in FIG. 6 for each pixel. When the encrypted image shown in FIG. 7 is superimposed on the random dot image 30 shown in FIG. 5, the first mask image 32 shown in FIG. 6 appears as shown in FIG. This is

1) can be indicated by: However, in (Equation 1), the random dot image 30 is expressed as a variable “変数”, the encrypted image 31 is expressed as a variable “Ε”, and the first mask image 32 is expressed as a variable “M”.

(Equation 1)

Overlapping the random dot image 30 with the encrypted image 31 can be represented by the logical sum of the variable 'K' and the variable 'E'. From (Equation 1), the logical OR of variables 'Κ' and 'E' is equal to the logical OR of variables 'Κ' and 'Μ'. In other words, it can be seen that this is the same as the superimposition of the random dot image 30 and the first mask image 32, so the superimposition of the random dot image 30 and the encrypted image 31 results in the first mask image 32. When it appears I understand that. Further, the encrypted image shown in FIG. 7 is read by a scanner, and the read encrypted image is subjected to an exclusive OR operation with the random dot image 30 shown in FIG. 5 to obtain the first mask image 3 shown in FIG. 2 can be completely restored. This can be shown by the following (Equation 2). From (Equation 2), the exclusive OR of the variable 'K' of the random dot image 30 and the variable 'E' of the cryptographic image 31 becomes equal to the variable 'Μ' of the first mask image 32. You can see that.

(Equation 2)

The geometric transformation performed in S15 will be described with reference to FIG. FIG. 9 schematically shows the mechanism of double-sided printing in the printer while indicating the paper feed path 41 indicated by the arrow of the paper 40 and the processing stages ST1 to ST3. In general, when performing double-sided printing, the paper is discharged while transporting a predetermined path while printing on the surface of the paper 40 (ST 1), and the paper is fed in the reverse direction via the paper feed roller 42. Then, the paper 40 is reversed (ST2, ST3), and printing is performed on the back side, so the path for printing the front side and the path for printing the back side are different. For this reason, there is a slight difference in the paper feeding accuracy, and the size in the paper feeding direction is slightly different between the front and the back. Further, the printing surface may be inclined depending on the mounting accuracy of the printing drum 43. Further, the printing position may be shifted due to a difference in the timing of starting printing on the front side and the back side. In general printers, these are treated as manufacturing errors, and are corrected to some extent but not completely. Since such a problem occurs when the yarns are joined together, there is no guarantee that printing is performed at exactly the same place on the front and back surfaces.

Here, the scaling of the image due to the difference in paper feed between the front and the back can be regarded as a linear transformation, and the transformation combining the scaling, rotation, and translation of the image is also a linear transformation. In order to cancel the manufacturing error of the printer, inverse transformations are performed respectively, but the inverse transformations of scaling, rotation, and translation are also linear transformations of scaling, rotation, and translation. Furthermore, when the printed paper 40 is viewed through the light, the random dot image printed on the back side must look at the same position as the symbol image on the front side. The force S required to convert the random dot image into a mirror image is also a linear conversion. From this, in S15, mirror image transformation, scaling transformation, rotation transformation, and translation transformation can be performed by performing only one geometric transformation. Specifically, this geometric transformation can be performed as in (Equation 3).

p '=: Ap + c ··· (Equation 3)

In (Equation 3), the coordinates before the conversion are set as a variable “p”, the coordinates after the conversion are set as a variable “p ′”, the conversion matrix is set as a variable “A”, and a constant “c” is used. The transformation matrix “A” and the constant 'c, are obtained in advance according to the characteristics of the printer 1. This makes it possible to accurately align the encrypted image on the front side and the random dot image on the back side and output a printout, so that the first mask image 32 can be viewed by double-sided printing. Become. Conversely, with a normal printer or copier, the images do not completely match on the front and back, so that even if two-sided copying is performed, the first mask image 32 will not be visible, preventing illegal copying. be able to.

The density conversion performed in S16 will be described with reference to FIGS. 1OA and 1OB. When observing an image from the front side by transmitting light through paper, the image on the back side looks faint. For this reason, if the random dot image and the encrypted image are printed at the same density on the front and the back, the density difference appears when the light is transmitted, so that the first mask image 32 becomes inconspicuous. In the present embodiment, in S16, as shown in FIGS. 10A and 10B, the encrypted image to be printed on the front side is converted into a pseudo-shade pattern such that it is thinner than the random dot image to be printed on the back side. For the sake of illustration, shades are represented by diagonal lines, but the pseudo shade pattern may be a halftone dot.

Further, the embedding of the second mask image 33 by the density conversion performed in S16 will be described with reference to FIG. Referring to the second mask image 33, the points corresponding to the white pixels of the second mask image 33 in the random dot image or the encrypted image are converted into the pseudo gray pattern 50, and the points corresponding to the black pixels are obtained. Is converted to a pseudo shade pattern 51. If the pseudo shade patterns 50 and 51 in Fig. 11 look the same in density, the pseudo pattern 50 is designed as a fine pattern, and the pseudo pattern 51 is designed as a coarse pattern. The image 33 is invisible and the second mask image 33 becomes visible only after copying. (Application example)

FIGS. 12A and 12B and FIGS. 13 and 13B show application examples according to the present embodiment. In FIGS. 12A and 12B, the encrypted image 31 is printed as a background image over the entire printing area on the front side of the paper, and the random dot image 30 is printed as a background image on the back side. In FIGS. 13A and 13B, the encrypted image 31 or the random dot image 30 is not printed on the entire print area, but is printed on a partial area. The symbol image 31 or the random dot image 30 may have some latent image, that is, any image may be embedded (combined).

(Effects of Embodiment)

As described above, according to the embodiment of the present invention, the first mask image 32 can be embedded in the printed matter. This is proof that it is genuine because it cannot be recovered without knowing the key code and cannot be forged. If the same key code is continuously used, there is a risk that the encrypted image 31 will be decrypted by a third party, but this risk can be reduced by using the identifier in combination with the identifier. In addition, by printing the random dot image 30 and the encrypted image 31 on both sides, a visible watermark can be realized with a normal copying machine (scanner + printer) or printer. Since the watermark by double-sided printing cannot be realized without accurate alignment, it is difficult for a third party to illegally copy and reproduce the watermark with a double-sided copying machine, thereby preventing illegal copying. On the other hand, in the technology disclosed in Japanese Patent Application Laid-Open No. 2002-142502, a show-through pattern appears with the shift of halftone dots to be combined. Forgery is easy because the show-through pattern appears even if the image is printed with a slight offset.

In addition, by selecting a pseudo shade pattern corresponding to the second mask image 33, it is possible to visually recognize the second mask image 33 that appears to rise when the image of the copy result is confirmed. Become. In other words, the real thing that has not been copied can be seen from the first mask image 32 that appears as a “watermark”, and when copied, the second mask image 33 appears and the copy (forgery) appears. At the same time) It becomes possible. On the other hand, the technique disclosed in the conventional Japanese Patent Application Laid-Open No. 54-74125 only provides a function in which a pattern appears when a printed matter is copied. The embodiments disclosed this time are to be considered in all respects as illustrative and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

Claims

The scope of the claims

1. An image processing device (1) for processing an image to be output,

First random dot image generating means (17) for generating a first random dot image in which points are randomly arranged;

A second random dot image generation for generating a second random dot image by performing a predetermined operation for each pixel corresponding to the generated first random dot image and a previously prepared first mask image Means (18),

The predetermined operation is performed through the first random dot image when the generated first random dot image and the second random dot image are superimposed and output so that the pixels match. The image processing device (1), characterized in that the calculation is an operation for visually appearing the first mask image.

2. It is further provided with an overlay processing means (19),

The superposition processing means (19) is configured such that when the generated first random dot image and the second random dot image are superimposed on each other and output, the corresponding pixels match. The image processing apparatus (1) according to claim 1, wherein the processing is performed as follows.

3. When printing an image on paper, the first random dot image is printed out on one side of the paper, and the second random dot image is printed out on the other side of the same paper. The image processing device (1) according to claim 2, wherein

4. The superposition processing means (19) includes a mirror image conversion means (20),

The image processing device according to claim 3, wherein the mirror image conversion means (20) converts one of the first random dot image and the second random dot image into a mirror image with respect to the other. (1).

5. The superposition processing means (19) includes an image conversion means (21),

The image conversion means (21) performs at least one of translational conversion, rotation conversion, and scaling conversion on at least one of the first random dot image and the second random dot image. An image processing device (1) according to claim 3.

6. The method according to claim 1, wherein the predetermined operation is an exclusive OR operation. Image processing device (1).

7. The method according to claim 6, wherein the first random dot image generating means (17) generates the first random dot image such that the appearance probability of a point is approximately 50%. Image processing device (1).

8. The image according to claim 1, wherein the first random dot image generating means (17) generates the first random dot image according to a random number sequence uniquely determined using key data. Processing equipment (1).

9. The image processing apparatus according to claim 8, wherein the key data includes a different identifier generated upon outputting an image.

10. The image processing device according to claim 9, wherein the identifier is output together with the second random dot image.

1 1. In addition, an image embedding means (22)

The image embedding means (22) stores each point of at least one of the first random dot image and the second random dot image at a different density according to a second mask image prepared in advance. 2 .The image processing according to claim 1, wherein a process of embedding the second mask image in at least one of the first and second random dot images is performed by converting the image into a pattern having the same density. 3. Equipment (1).

12. When printing out an image on paper, it is assumed that the first random dot image is printed out on one side of the paper and the second random dot image is printed out on the other side of the same paper. The image processing device (1) according to claim 1, characterized in that:

13. An image density conversion means (23) is further provided.

The image density conversion means (23) is configured to at least change the image density of the first random dot image and the second random dot image according to the light transmittance of the printing surface of the paper so that the image density differs. The image processing device (1) according to claim 12, wherein the image processing device converts the density of one image.

14. An image processing method for processing an image to be output,

A first random dot image generating step of generating a first random dot image in which points are randomly arranged;

The generated first random dot image and a previously prepared first mask image A second random dot image generating step of generating a second random dot image by performing a predetermined operation for each pixel corresponding to

The predetermined operation is performed through the first random dot image when the generated first random dot image and the second random dot image are superimposed and output so that the pixels match. An image processing method for visually causing the first mask image to appear.

15. An image processing program for causing a computer to execute an image processing method for processing an image to be output,

The image processing method includes:

A second random dot image generation for generating a second random dot image by performing a predetermined operation for each pixel corresponding to the generated first random dot image and a previously prepared first mask image With steps

The predetermined operation is performed through the first random dot image when the generated first random dot image and the second random dot image are superimposed and output so that the pixels match. An image processing program for causing a computer to execute an image processing method, wherein the image processing method is an operation for causing the first mask image to appear visually.

16. A machine-readable recording medium recording an image processing program for causing a computer to execute an image processing method for processing an image to be output,

The image processing method includes:

The predetermined operation is performed when the generated first random dot image and the second random dot image are output in such a manner that pixels are superimposed on each other so as to coincide with each other. A machine-readable recording medium on which an image processing program is recorded, the computation being an operation for visually appearing the first mask image via the first random dot image.