JP4446185B2 - Information generating apparatus and method, and program - Google Patents

Description

  The present invention relates to an information generation apparatus, method, and program, and is suitably applied to, for example, guaranteeing that a sheet with a predetermined content is an original.

  Conventionally, paper has been used as an object for printing various contents such as documents and photographs, and photographic paper on which the contents (hereinafter referred to as “print contents”) are printed is, for example, commodity exchange media such as money, certificates, etc. It often has high value because it functions as various media such as content certification media or information storage media such as personal works.

For this reason, various countermeasures for preventing unauthorized duplication of the print contents printed on the photographic paper are considered, and as such a countermeasure, the pattern information obtained based on the pattern on the photographic paper is acquired. An unauthorized duplication prevention device that protects the contents printed on photographic paper by storing the information on the photographic paper and verifying the legitimacy of the photographic paper based on the stored pattern information is disclosed by the present applicant. Proposed.
PCT / JP2005 / 001176

  By the way, in this unauthorized duplication preventing apparatus, the print content printed on the photographic paper is protected by verifying the legitimacy of the photographic paper based on the pattern information stored on the photographic paper. In addition to the pattern information stored on the photographic paper, if it can be assured that the photographic paper is the original by the pattern information not stored on the photographic paper, the print contents printed on the photographic paper will be more appropriately protected. Can be considered.

  The present invention has been made in consideration of the above points, and an object of the present invention is to propose an information generation apparatus, method, and program capable of appropriately protecting the contents attached to paper.

  In order to solve such a problem, in the present invention, an image pickup unit that picks up an image of a unique pattern on a paper with a predetermined content, and a first extraction unit that extracts a feature of the pattern image picked up by the image pickup unit, , Second extraction means for extracting features from the pattern image in more detail than the first feature extracted by the first extraction means, storage means for storing the first features on the paper, and second The second feature extracted by the extraction means is provided with a registration means for registering in the predetermined storage means.

  Therefore, when simplicity or speed is the first priority, it is possible to easily and quickly verify that the paper is different from the original by collating with the first feature. In the case where the priority is given to the first characteristic, the second feature representing the detailed feature amount is extracted as compared with the first feature. It is possible to verify precisely and rigorously that it is different.

  In addition, a first step of imaging a unique pattern on a paper with a predetermined content, a feature of the pattern image captured in the first step is extracted as a first feature, and a first in the pattern image is extracted. A second step of extracting a feature more detailed than the feature of the second feature as a second feature, and a third step of storing the first feature on paper and registering the second feature in a predetermined storage means I made it.

  Therefore, when simplicity or speed is the first priority, it is possible to easily and quickly verify that the paper is different from the original by collating with the first feature. In the case where the priority is given to the first characteristic, the second feature representing the detailed feature amount is extracted as compared with the first feature. It is possible to verify precisely and rigorously that it is different.

  Furthermore, the program extracts a first image of the unique pattern on the paper with a predetermined content and the feature of the pattern image captured in the first step as the first feature, and the pattern image. A second process for extracting a feature more detailed than the first feature as a second feature, and storing the first feature on paper and registering the second feature in a predetermined storage means The processing was executed.

  Therefore, when simplicity or speed is the first priority, it is possible to easily and quickly verify that the paper is different from the original by collating with the first feature. In the case where the priority is given to the first characteristic, the second feature representing the detailed feature amount is extracted as compared with the first feature. It is possible to verify precisely and rigorously that it is different.

  According to the present invention, an image of a unique pattern on a paper with a predetermined content is captured, and the feature of the pattern image captured in the first step is extracted as a first feature, and the first feature in the pattern image is extracted. By extracting a feature more detailed than the feature as the second feature, storing the first feature on paper, and registering the second feature in a predetermined storage means, it is easy or quick. Is the first priority, it is possible to easily and quickly verify that the paper is different from the original by collating with the first feature, and notarity is the first priority. In such a case, in order to extract the second feature representing a detailed feature amount compared to the first feature, it is possible to accurately determine that the paper is different from the original by comparing with the second feature. And can be rigorously verified and thus attached to the paper The contents can be realized suitably be protected information generating apparatus and method, and program.

  Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

(1) Original guarantee method As shown in FIG. 1, the paper has a unique pattern (hereinafter referred to as a pattern) constituted by complex intertwining of fibers inside the paper instead of the surface. Yes. This pattern can be obtained as an image (hereinafter referred to as a pattern image) by, for example, a transmissive scanner or the like, as can be seen from the fact that it can be visually confirmed by holding it over light.

  Therefore, in the original guarantee method according to the present embodiment, a pattern feature amount (hereinafter referred to as a pattern feature amount) based on a pattern image obtained from such paper is extracted, and printing is performed using the pattern feature amount. Guarantee that the paper is the original.

  In this case, in the original guarantee method according to the present embodiment, as shown in FIG. 2, a predetermined area (hereinafter referred to as “designated area”) of the original photographic paper (hereinafter referred to as “original photographic paper”) OP as shown in FIG. A first pattern feature amount MP1 is extracted from the pattern image MG of the pattern in the AR).

  In the original guarantee method in the present embodiment, the first pattern feature amount MP1 is a two-dimensional barcode (hereinafter referred to as the first pattern code) BC1 and a part of the photographic paper surface of the original photographic paper OP. Thus, the first pattern feature MP1 is stored in the original photographic paper OP.

  In the original guarantee method according to the present embodiment, a second feature amount that is more detailed than the first pattern feature amount MP1 described above is obtained from the pattern image MG of the pattern in the designated area AR of the original photographic paper OP. The pattern feature amount MP2 is extracted.

  In the original guarantee method according to the present embodiment, the second pattern feature value MP2 is extracted as described above using the two-dimensional barcode (hereinafter, referred to as a second pattern code) BC2. Is registered in the pattern code storage unit in association with the pattern code BC1.

  On the other hand, in the original guarantee method according to the present embodiment, when simplicity or speed is given first priority, as shown in FIG. 3, a photographic paper (first paper pattern BC1 printed) (see FIG. 3). In the case where XPc exists (hereinafter referred to as “code-attached photographic paper”), the first image described above is used from the pattern image (hereinafter referred to as “code-attached pattern image”) MGX in the designated area AR of this code-added photographic paper XPc. The first matching pattern feature value MPC1 is extracted by performing the same process as that for extracting the pattern feature value MP1.

  In the original guarantee method according to the present embodiment, the first matching pattern feature value MPC1 is compared with the first restored pattern feature value MPR1 restored based on the first pattern code BC1, and the code is attached. The legitimacy and validity of the photographic paper XPc, that is, whether or not the code-added photographic paper XPc is the original is verified.

  Here, in the original guarantee method according to the present embodiment, when a matching rate equal to or higher than a predetermined threshold is obtained as a result of this collation, it is determined that the code-added photographic paper XPc is a valid original photographic paper OP itself. This guarantees that the code-added photographic paper XPc is the original.

  On the other hand, in the original guarantee method according to the present embodiment, when a matching rate lower than a predetermined value is obtained as a collation result, the code-added photographic paper XPc is copied instead of the valid original photographic paper OP. It is determined that the code-added photographic paper XPc is the original.

  On the other hand, in the original guarantee method in the present embodiment, when notarity is given first priority, as shown in FIG. 4, the code-attached pattern image in the designated area AR of the code-added photographic paper XPc is used. The second matching pattern feature value MPC2 is extracted from the MGX by performing the same process as the extraction process of the second pattern feature value MP2 described above.

  In the original guarantee method according to the present embodiment, the pattern code storage unit searches for the first pattern code BC1, and is associated with the first pattern code BC1 registered in the pattern code storage unit. When the second pattern code BC2 is detected, the second restored pattern feature value MPR2 is restored based on the second pattern code BC2.

  In the original guarantee method according to the present embodiment, the second collation pattern feature amount MPC2 and the second restored pattern feature amount MPR2 are collated to determine whether or not the coded photographic paper XPc is the original document itself. Similar to the case where the first matching pattern feature value MPC1 and the first restored pattern feature value MPR1 are matched, a matching rate equal to or higher than a predetermined threshold is obtained as a matching result. In this case, it is ensured that the code-added photographic paper XPc is the original, but if a matching rate lower than a predetermined value is obtained as a collation result, it is confirmed that the code-added photographic paper XPc is the original. Deny.

  In this way, in the original guarantee method according to the present embodiment, as shown in FIG. 5, even if the print content on the original photographic paper OP is copied (Copyed) to the code-added photographic paper XPc, Since the same pattern image existing in the designated area AR of the original photographic paper OP cannot exist on the photographic paper XPc, the collation result does not match, and the code-added photographic paper XPc is a duplicate and not an original. It is made so that it can be verified reliably.

  Therefore, in the original guarantee method according to the present embodiment, when simplicity or speed is given first priority, the first matching pattern feature amount MPC1 and the first restored pattern feature amount MPR1 Can be verified easily and quickly that the code-added photographic paper XPc is different from the original, and the first pattern feature amount is given when notarity is given first priority. In order to extract the second pattern feature quantity MP2 representing a detailed feature quantity compared to MP1, the second matching pattern feature quantity MPC2 and the second restored pattern feature quantity MPR2 are collated to obtain the first pattern feature quantity MP2. The second matching pattern feature value MPC2 and the second restored pattern feature value MPR2 can be verified accurately and strictly.

(2) Configuration of Original Guarantee Device In FIG. 6, reference numeral 1 denotes an original guarantee device according to the present embodiment. The control unit 2 that controls the entire original guarantee device 1 is connected to the scanner unit 4 via the bus 3. The printer unit 5 and the pattern code storage unit 6 are connected to each other.

  The control unit 2 includes a central processing unit composed of a CPU (Central Processing Unit), a work memory composed of a RAM (Random Access Memory), and an information storage memory composed of a ROM (Read Only Memory).

  In this information storage memory, the position information (hereinafter referred to as area position information) of the designated area AR (FIG. 2) for each standard size of paper (hereinafter referred to as area position information), character string information (hereinafter referred to as this). Various information such as code character string information) and programs are stored. Therefore, the control unit 2 executes various processes according to programs loaded in the work memory, using various information stored in the information storage memory as appropriate.

  In practice, the control unit 2 prints the first pattern code BC1 (FIG. 2) and gives a predetermined command for registering the second pattern code BC2 (FIG. 2) from an operation unit (not shown). The pattern image reading command is sent to the scanner unit 4.

  Then, as a response result of this command, the control unit 2 is supplied with data of the pattern image MG (FIG. 2) on the original photographic paper OP (FIG. 2) (hereinafter referred to as pattern image data) D1 from the scanner unit 4. If this is the case, a transition is made to the first mode (hereinafter, this mode is referred to as a code print registration mode).

  In this case, the control unit 2 extracts the first pattern feature amount MP1 from the pattern image MG in the pattern image data D1, and also represents a second feature amount that is more detailed than the first pattern feature amount MP1. The pattern feature amount MP2 is extracted.

  Then, the control unit 2 sends the extracted first pattern feature amount MP1 to the printer unit 5 as two-dimensional barcode character string data (hereinafter referred to as first pattern code data) D2. As a result, the first pattern code data D2 is printed on the related original photographic paper OPR as the first pattern code BC1 (FIG. 2A) in the printer unit 5.

  Further, the control unit 2 extracts the extracted second pattern feature value MP2 as two-dimensional barcode character string data (hereinafter referred to as second pattern code data) D3 as described above. The first pattern code data D2 and the second pattern code data D3 are sent to the pattern code storage unit 6 in association with the first pattern code data D2. As a result, the first pattern code data D2 and the second pattern code data D3 are associated with the first pattern code BC1 and the second pattern code BC2, and the pattern for storing the pattern code is stored. It is registered in the code storage unit 6.

  On the other hand, when a predetermined command for simply or quickly verifying whether or not the code-added photographic paper XPc (FIG. 3) is an original is given from the operation unit (not shown), the control unit 2 An image reading command and a code reading command are sent to the scanner unit 4.

  Then, as a response result of these commands, the control unit 2 sends data of the code-attached pattern image MGX (FIG. 3) on the code-added photographic paper XPc from the scanner unit 4 (hereinafter referred to as code-attached pattern image data) D4. When the first pattern code data D1 that is the read result of the first pattern code BC1 (FIG. 3) printed on the code-added photographic paper XPc and the scanner unit 4 are given, the second mode ( Hereinafter, this mode is referred to as a first verification mode).

  In this case, the control unit 2 performs the same process as the extraction process of the first pattern feature quantity MP1 described above from the code pattern image MGX in the code pattern image data D4, thereby performing the first matching pattern feature quantity. MPC1 is extracted.

  Then, the control unit 2 compares the first matching pattern feature quantity MPC1 with the first restored pattern feature quantity MGR1 based on the first pattern code data D1, and as a result of the comparison, each feature quantity is It is determined whether or not they coincide with each other beyond a predetermined threshold, and only when it can be recognized that they coincide, it is proved that the code-added photographic paper XPc is the original photographic paper OP itself, that is, the original. Has been made.

  On the other hand, when a predetermined command for verifying whether or not the code-added photographic paper XPc (FIG. 3) is the original is given from the operation unit (not shown), the control unit 2 reads the pattern image reading command. The code reading command is sent to the scanner unit 4.

  Then, as a response result of these commands, the control unit 2 outputs the code-attached pattern image data D4 on the code-added photographic paper XPc from the scanner unit 4 and the first pattern code BC1 printed on the code-added photographic paper XPc (see FIG. When the first pattern code data D1 which is the reading result of 3) is given, the mode transits to a third mode (hereinafter, this mode is referred to as a second verification mode).

  In this case, the control unit 2 performs the same processing as the extraction processing of the second pattern feature amount MP2 described above from the code pattern image MGX in the code pattern image data D4, thereby performing the second matching pattern feature amount. MPC2 is extracted.

  In addition, the control unit 2 searches the pattern code storage unit 6 for the first pattern code BC1 in the supplied first pattern code data D1 to thereby register the first pattern code stored in the pattern code storage unit. When the second pattern code BC2 associated with the pattern code BC1 is detected, the second pattern code BC2 is used as the second pattern code data D3 from the pattern code storage unit 6 to the control unit 2. Read.

  Then, the control unit 2 collates the second matching pattern feature value MPC2 with the first restored pattern feature value MGR1 based on the first pattern code data D1. It is determined whether or not they coincide with each other beyond a predetermined threshold value, and only when it can be recognized that both coincide with each other, it is proved that the code-added photographic paper XPc is the original.

  In this way, in the code print registration mode, the control unit 2 prints the first pattern feature amount MP1 based on the pattern image MG on the original photographic paper OP as the first pattern code BC1 on the original photographic paper OP and the pattern. The second pattern feature amount MP2 based on the image MG is registered in the pattern code storage unit 6 as the second pattern code BC2 in association with the first pattern code BC1, and is printed in the first verification mode. Only when the first restored pattern feature value MGR1 based on the first design code BC1 and the first design feature value MP1 match, it is guaranteed to be an original and is registered in the second verification mode. Only when the second restored pattern feature value MGR2 based on the second design code BC2 and the second pattern feature value MP2 coincide with each other is the original. It has been made to ensure that.

  On the other hand, the scanner unit 4 has a transmission mode, a reflection mode, and a code reading mode. When a pattern image reading command is given from the control unit 2, the scanning mode is given, or when a predetermined duplication permission command is given. Is configured to execute a reflection mode and a code reading mode when a code reading command is given.

  In practice, in the transmission mode, the scanner unit 4 irradiates the original photographic paper OP or the code-added photographic paper XPc placed on the document table with light, and optically transmits the pattern projection light obtained by transmitting the light. An image is formed on the solid-state imaging device through the system. Then, the scanner unit 4 performs A / D (Analog / Digital) conversion processing or the like on the pattern image signal obtained from the solid-state image sensor, and the pattern image data D1 and D2 or code pattern image data obtained as a result. D4 is sent to the control unit 2.

  In the reflection mode, the scanner unit 4 emits light to the original printing paper OP placed on the document table, for example, and reflects the print content reflection light obtained by reflecting the light through the optical system. Form an image on the element. The scanner unit 4 performs A / D conversion processing or the like on the print content image signal obtained from the solid-state imaging device, and sends the print content image data D5 obtained as a result to the printer unit 5.

  Further, in the code reading mode, the scanner unit 4 activates the two-dimensional code reader 4a connected to the scanner unit 4 and reads the first pattern code data D2 supplied by being read by the two-dimensional code reader 4a. It is sent to the control unit 2.

  In this way, the scanner unit 4 can read the pattern image, the pattern code, or the print content by executing the mode according to the various commands given from the control unit 2.

  On the other hand, the printer unit 5 uses two-dimensional code font information (hereinafter referred to as code font information) and position information of the first pattern code BC1 for various standard sizes of paper (hereinafter referred to as code position information). Are stored in the internal memory, and the printing process is executed using these pieces of information as appropriate.

  In this case, when the first pattern code data D2 is given from the control unit 2, the printer unit 5 performs a pulse width modulation process or the like on the first pattern code data D2, and print data obtained as a result. Is sent to the print head. As a result, by driving the print head unit based on the print data, code font information, and code position information, the first pattern code BC1 is placed at a predetermined position on the print paper (original print paper OP) set on the print paper stand at this time. Will be printed.

  Further, when the print content image data D5 is given from the scanner unit 4, the printer unit 5 performs a pulse width modulation process or the like on the print content image data D5, and sends the resulting print data to the print head unit. To do. As a result, by driving the print head unit based on the print data, the print contents of the original print paper OP are duplicated on the paper set on the print paper stand at this time.

  In this way, the printer unit 5 can print the first pattern code BC1 based on the first pattern code data D2 supplied from the control unit 2, and can duplicate the print content based on the print content image data D5. It has been made possible.

(3) Processing in Control Unit (3-1) Processing in Control Unit in First Embodiment Here, code print registration mode and first verification mode in control unit 2 in the first embodiment. When the processing contents in the second verification mode are functionally classified, as shown in FIG. 7, the low-frequency component extraction unit 10, the image separation unit 11, the region division unit 12, and the pattern feature extraction The unit 13, the two-dimensional code conversion unit 14, and the collation unit 15 can be divided.

  In this case, in the code print registration mode, the control unit 2 performs the low frequency component extraction unit 10, the image separation unit 11, the region division unit 12, and the pattern feature amount with respect to the pattern image data D 1 given from the scanner unit 4. Various processes are sequentially performed through the extraction unit 13 and the two-dimensional code conversion unit 14, and the first pattern code data D2 obtained as a result is sent to the printer unit 5, and the first pattern code data D2 and the second pattern code data D2 The pattern code data D3 is sent to the pattern code storage unit 6 in a state of being associated therewith.

  In the first verification mode, the control unit 2 applies the low frequency component extraction unit 10, the image separation unit 11, the region division unit 12, and the pattern feature to the code-attached pattern image data D4 given from the scanner unit 4. After performing various processes sequentially through the quantity extraction unit 13, the collation unit 15 performs collation processing based on the processing result and the first pattern code data D2 given from the scanner unit 4.

  Further, in the second verification mode, the control unit 2 sequentially passes the low-frequency component extraction unit 10, the image separation unit 11, and the region division unit 12 with respect to the code-pattern image data D4 provided from the scanner unit 4. After performing various processes, the collation unit 15 performs collation processing based on the processing result and the second pattern code data D3 given from the pattern code storage unit 6.

  Hereinafter, low-frequency component extraction processing by the low-frequency component extraction unit 10, image separation processing by the image separation unit 11, image division processing by the region division unit 12, pattern pattern extraction processing by the pattern pattern extraction unit 13, two-dimensional The two-dimensional code conversion process by the code conversion unit 14 and the verification process by the verification unit 15 will be described in detail.

(3-1-1) Low-frequency component extraction process The low-frequency component extractor 10 is in the code print registration mode, as shown in FIG. 8, the pattern image MG (FIG. 8) on the original photographic paper OP (FIG. 2). From (A)), a low-frequency component image (hereinafter referred to as a low-frequency pattern image) MGL (FIG. 8B) is extracted.

  Specifically, the low frequency component extraction unit 10 performs a Fourier transform process on the pattern image MG in the pattern image data D1 provided from the scanner unit 4 based on the region position information stored in the internal memory. Generate component data.

  Subsequently, the low frequency component extraction unit 10 sets the data value of the high frequency component equal to or higher than a predetermined threshold to “0” for the frequency component data, and then performs an inverse Fourier transform process to thereby generate the low frequency pattern image MGL. Data (hereinafter referred to as low-frequency pattern image data) D10 (FIG. 7) is generated and sent to the image separation unit 11.

  In this way, the low frequency component extraction unit 10 extracts the low frequency pattern image MGL, thereby removing various noise components generally included in the high frequency components of the image, such as noise of the solid-state imaging device in the scanner unit 4, for example. Has been made to be able to.

  As a result, the low frequency component extraction unit 10 can avoid a decrease in the extraction accuracy of the pattern feature quantity in the pattern feature quantity extraction unit 13 due to various noise components. The reliability in the collation result of the collation process can be improved.

  On the other hand, in the first verification mode and the second verification mode, the low frequency component extraction unit 10 uses the pattern image MGX (FIG. 8A) on the code-added photographic paper XPc (FIGS. 3 and 4) as described above. By performing the same extraction process as that of the pattern image MG of, the low frequency component pattern image (hereinafter referred to as a coded low frequency pattern image) MGXL (FIG. 8B) data (hereinafter referred to as this). D11 (referred to as low-frequency pattern image data with code) (FIG. 7) is generated and transmitted to the image separation unit 11.

(3-1-2) Image Separation Processing As shown in FIG. 9, the image separation unit 11 performs the low-frequency pattern image MGL extracted by the low-frequency component extraction unit 10 or the coded low-frequency pattern image MGXL (FIG. 9). (A)) is separated into a white component pattern image WIM (FIG. 9B) and a black component pattern image BIM (FIG. 9C).

  Specifically, the image separation unit 11 includes the low frequency pattern image MGL of the low frequency pattern image data D10 supplied from the low frequency component extraction unit 10 or the low frequency pattern image MGXL with code of the low frequency pattern image data D11 with code. The luminance value is sequentially detected for each pixel, and pixels other than the pixel (hereinafter referred to as a white pixel) whose detection result is a luminance value equal to or lower than a predetermined low luminance threshold (hereinafter referred to as a white threshold) are detected. After extracting an image of a low luminance component (hereinafter referred to as a white component pattern image) MGW (FIG. 9B) so as to be converted to the highest luminance level, this white component pattern image MGW is converted into data (hereinafter referred to as “white component pattern image”). This is called white component pattern image data) and is sent to the area dividing unit 12 as D12.

  Further, the image separation unit 11 has a luminance value at which the detection result of the luminance value of each pixel in the low frequency pattern image MGL or the coded low frequency pattern image MGXL is equal to or higher than a predetermined high luminance threshold value (hereinafter referred to as a black threshold value). An image having a high luminance component (hereinafter referred to as a black component pattern image) MGB (FIG. 9 (C )) Is extracted, and this black component pattern image MGB is sent to the area dividing unit 12 as data (hereinafter referred to as black component pattern image data) D13.

  In this way, the image separation unit 11 can reduce the degree of complexity of the pattern by separating the white component pattern image MGW and the black component pattern image MGB.

  As a result, the image separation unit 11 can avoid a decrease in the extraction accuracy of the pattern feature quantity in the pattern feature quantity extraction unit 13 due to the high degree of complexity. The reliability of the collation result of the collation process can be improved.

(3-1-3) Area Division Processing As shown in FIG. 9, the area division unit 12 sets a set of adjacent white pixels to a pattern having a white component pattern image MGW (FIG. 9B). The area is divided into areas (hereinafter referred to as white damashes) as units, and the pattern included in the black component pattern image MGB (FIG. 9C) is defined as an area (hereinafter referred to as a group of adjacent black pixels). This is called black dama).

  Specifically, after the area dividing unit 12 detects all white pixels from the white component pattern image MGW of the white component pattern image data D12 supplied from the image separation unit 11, as shown in FIG. A total of 8 pixels (hereinafter referred to as 8 neighboring pixels) of 4 pixels in the vertical and horizontal directions and 4 pixels in the oblique direction adjacent to an arbitrary target pixel AP are sequentially connected.

Then, for example, as shown in FIG. 10B, the area dividing unit 12 associates the identification information with the group of white pixels that have been connected so far when the white pixel is no longer detected in the eight neighboring pixels. ₁ , WD ₂ ..., And WD _n .

The area dividing unit 12 also applies a plurality of black dams BD (BD _{1 to} BD _{n) to} the black component pattern image MGB of the black component pattern image data D13 supplied from the image separation unit 11 in the same manner as the white component pattern image MGW. ).

In this way, the area dividing unit 12 divides the pattern included in the white component pattern image MGW into a plurality of white damashes WD (WD _{1 to} WD _n ), and the pattern included in the black component pattern image MGB. By dividing into BD (BD _{1 to} BD _n ), the pattern can be subdivided.

  As a result, the area dividing unit 12 can finely analyze the patterns included in the white component pattern image MGW and the black component pattern image MGB. Therefore, the pattern feature amount extraction accuracy of the pattern feature amount extraction unit 13 is improved. As a result, the reliability of the collation result of the collation process in the collation unit 15 can be improved.

In addition to such a configuration, the area dividing unit 12 divides the pattern included in the white component pattern image MGW into a plurality of white dama WDs (WD _{1 to} WD _n ), as shown in FIG. 11A, for example. As shown in FIG. 11 (B), a dama that is equal to or less than a predetermined number of connections (hereinafter referred to as a small dama) is removed from each white dama WD, and the result obtained by the removal is obtained. The white dama WD (WD _{1 to} WD _n ) is sent to the pattern feature quantity extraction unit 13 as data (hereinafter referred to as white dama data) D14 (FIG. 7).

The regions separate portions 12, black lumps BD _(BD 1 ~BD _n) in the same manner as white lumps WD _(WD 1 ~WD _n to remove small lumps also a result of the removal of the resulting black lumps BD ( BD _{1 to} BD _n ) are sent to the pattern feature quantity extraction unit 13 as data (hereinafter referred to as black dama data) D15 (FIG. 7).

  As a result, the area dividing unit 12 can extract only the feature portions of the pattern included in the white component pattern image MGW and the black component pattern image MGB as white dama WD and black dama BD. The extraction accuracy of the pattern feature amount can be further improved.

Further, in the second verification mode, the area dividing unit 12 sets the white dama WD (WD _{1 to} WD _n and black dama BD (BD _{1 to} BD _n ) as the second matching pattern feature amount MPC2, and the white dama data D14. The black dama data D15 is sent to the collation unit 15 as second collation pattern feature value MPC2 data (hereinafter referred to as second collation pattern feature value data) D16.

(3-1-4) Pattern Feature Amount Extraction Processing The pattern feature amount extraction unit 13 uses each white dama WD (WD _{1 to} WD _n ) and each black dama BD (BD _{1 to} BD _n ) in the code print registration mode. The first pattern feature quantity MP1 and the second pattern feature quantity MP2 are extracted by calculating the shape feature quantity.

In this case, the pattern feature amount extraction unit 13 approximates each of the white dama WD and the black dama BD to a rectangle in order to easily and quickly extract the feature amounts of the white dama WD and the black dama BD. That is, as shown in FIG. 12, the pattern feature quantity extraction unit 13 performs center coordinates (x _c , y _c ), long side l, short side w, and long length for each dama (white dama WD or black dama BD). An angle θ formed by the side l and the horizontal axis (hereinafter referred to as a rectangular information value) is calculated as a feature amount.

Specifically, the pattern feature quantity extraction unit 13 calculates individual feature quantities for each white dama WD (WD _{1 to} WD _n ) of the white dama data D14 supplied from the area division unit 12. If the luminance value of the pixels constituting the white dama WD is I (x, y),

The primary image moment M ₀₀ , the secondary image moments M ₁₀ , M ₀₁ and the tertiary image moments M ₂₀ , M ₀₂ , M ₁₁ are calculated according to the image moment M _pq defined by

Then, the pattern feature quantity extraction unit 13 uses these primary, secondary, and tertiary image moments M ₀₀ , M ₁₀ , M ₀₁ , M ₂₀ , M ₀₂ , and M ₁₁ to use the center coordinates (x _c , y _c ). The following formula

And the long side l and the short side w

And calculate the angle θ

Calculate according to

In this way, the pattern feature quantity extraction unit 13 calculates the feature quantity (rectangular information value) for each white dama WD (WD _{1 to} WD _n ).

The pattern feature quantity extraction unit 13 also applies the black dama BD (BD _{1 to} BD _n ) of the black dama data D15 supplied from the area dividing unit 12 in the same manner as the white dama WD (WD _{1 to} WD _n ). The feature amount (rectangular information value) for each black dama BD is calculated using equations (1) to (4).

Thus calculated white lumps WD _(WD 1 ~WD _n) and the black lumps BD _(BD 1 ~BD _n) each feature quantity (rectangle information values) A pattern image MG of A pattern having a (FIG. 8 (A)) Since the value represents a characteristic shape, it means the extraction result itself of the first pattern feature amount MP1 included in the pattern image MG.

  Then, the pattern feature quantity extraction unit 13 sends the data as the first pattern feature quantity MP1 data (hereinafter referred to as first pattern feature quantity data) D17 (FIG. 7) to the two-dimensional code conversion unit. It is made like that.

  In this way, the pattern feature quantity extraction unit 13 approximates each white dama WD and black dama BD to a rectangle (hereinafter referred to as a rectangle) in order to easily and quickly extract the feature quantities of the white dama WD and black dama BD. The first pattern feature quantity MP1 is extracted based on the first extraction method).

Further, the pattern feature quantity extraction unit 13 extracts each white dama WD (WD _{1 to} WD _n ) and each black dama BD (BD) in order to accurately and rigorously extract the feature quantities of the white dama WD and the black dama BD. _{1 to} BD _n ), a point for generating a Bezier curve based on a point on the outer periphery of the dama (hereinafter referred to as a dama outer peripheral point) (hereinafter referred to as a control point). ) A sequence is determined, and these control point sequences are extracted as feature quantities.

Specifically, the pattern feature quantity extraction unit 13 uses the white dama data WD (WD _{1 to} WD _n ) of the white dama data D14 and the black dama data BD (BD _{1 to} BD _n ) of the black dama data D15 supplied from the area division unit 12. An area (hereinafter referred to as a dama total area) is calculated based on the number of pixels, and a correspondence table between a dama total area (number of pixels) stored in advance in the internal memory, the lattice size of a square lattice, and the order of a Bezier curve Referring to FIG. 5, the grid size and the order of the Bezier curve corresponding to the number of pixels detected at this time are switched.

  Then, the pattern feature quantity extraction unit 13 divides the designated area AR of the white dama data D14 and the designated area AR of the black dama data D15 by the square grid of the grid size switched at this time, and generates the n-th order Bezier curve switched at this time. A control point sequence composed of “n + 1” control points is determined for each of the white dama WD and the black dama BD in the designated area AR.

Here, a method for determining such a control point sequence will be specifically described. For convenience of explanation, here, a control point sequence including four control points for generating a cubic Bezier curve for the white dama WD ₁ is described. The case of determining will be described with reference to FIG.

In FIG. 13, the pattern feature quantity extraction unit 13 recognizes the intersection points P _{1 to} P ₁₂ between the square lattice and the outer periphery of the white dama WD ₁ as control points. For example, the pattern feature quantity extraction unit 13 uses the control point P ₁ as the first start point. The four control points to be performed are sequentially selected as control point sequences P _{1 to} P ₄ , P _{4 to} P ₇ , P _{7 to} P ₁₀ , and P _{10 to} P ₁₂ .

In this case, the pattern feature quantity extraction unit 13 includes the end points (control points P ₄ , P ₇ , P ₁₂₎ of the control point sequences P _{1 to} P ₄ , P _{4 to} P ₇ , P _{7 to} P ₁₀ , and P _{10 to} P _{12. 10)} start of the next control points (control points _{P 4,} P _{7, P 10)} being adapted to select as, also the last control points _P 10 _{to P} as ₁₂ in the three control points The control point sequence is selected.

Here, when these control point sequences P _{1 to} P ₄ , P _{4 to} P ₇ , P _{7 to} P ₁₀ , and P _{10 to} P ₁₂ are determined as they are as control point sequences for the white dama WD ₁ , Bezier curve generated from P ₁ to P _12, since the respective inner or outer than the outer periphery of the white lumps WD _1, so that the actual white lumps WD ₁ thus obtained as extremely different lumps.

Therefore, in the case of this embodiment, the pattern feature amount extraction unit 13 controls the control points between the start points and the end points in the control point sequences P _{1 to} P ₄ , P _{4 to} P ₇ , and P _{7 to} P ₁₀ (hereinafter referred to as this). (Referred to as intermediate control points) P ₂ and P ₃ , P ₅ and P ₆ , P ₈ and P ₉ , the last control point sequence P _{10 to} P ₁₂ and the first control point sequence P _{1 to} P ₄ (P ₁₀ , P ₁ ) (hereinafter also referred to as intermediate control points) P ₁₁ and P ₁₂ are shifted inward or outward from the outer periphery of the white dama WD ₁ .

Specifically, when the middle control points P ₂ and P ₃ are set to the center of point symmetry, the pattern feature amount extraction unit 13 generates a perpendicular to the line segment P ₁ -P ₄ from the middle control points P ₂ and P _3. Points C ₂ and C ₃ corresponding to the intersections Q ₂ and Q ₃ with the line segment P ₁ -P ₄ are detected, and the detected points C ₂ and C ₃ and the control points P ₁ and P ₄ are connected to the control point sequence. Determine as P ₁ -C ₂ -C ₃ -P ₄ .

Then, the pattern feature quantity extraction unit 13 also applies the points C ₅ and P ₅ and P ₆ , P ₈ and P ₉ , P ₁₁ and P ₁₂ to the points C ₅ and P _{3 in the} same manner as the intermediate control points P ₂ and P _3. C ₆ , C ₈ and C ₉ , C ₁₁ and C ₁₂ are detected respectively, and the points C ₅ and C ₆ , C ₈ and C ₉ , C ₁₁ and C _12, and the corresponding control points P ₄ and P ₇ , P ₇ and P ₁₀ , P ₁₀ and P ₁₂ are respectively designated as control point sequences P ₄ -C ₅ -C ₆ -P ₇ , P ₇ -C ₈ -C ₉ -P ₁₀ , P ₁₀ -C ₁₁ -C _12. decide.

Therefore, the pattern feature quantity extraction unit 13 controls the control point sequence P ₁ -C ₂ -C ₃ -P ₄ , P ₄ -C ₅ -C ₆ -P ₇ , P ₇ -C ₈ -C for the white dama WD _{1. 9 -P 10, P 10 -C 11} -C 12 a is adapted to determine, respectively.

As described above, the pattern feature quantity extraction unit 13 selects the intermediate control point sequence among the selected control point sequences (P _{1 to} P ₄ , P _{4 to} P ₇ , P _{7 to} P ₁₀ , P _{10 to} P ₁₂ ). by shifting the control point _{(P 2} and _P 3, _{P 5} and _P 6, _{P 8} and _{P 9, P 11} and _{P 12)} inside or outside the outer periphery of the white lumps WD _1, the actual white lumps WD ₁ Control point sequences (P ₁ -C ₂ -C ₃ -P ₄ , P ₄ -C ₅ -C ₆ -P ₇ , P ₇ -C ₈ -C ₉ -P ₁₀ , P ₁₀ -C ₁₁ -C ₁₂ ) can be extracted as a pattern feature quantity.

The A pattern feature quantity extraction unit 13, white lumps _WD 2 _{~WD n} and for even the black lumps BD black Damadeta D14 supplied from the area separate portions 12 _(BD 1 ~BD _n), white lumps WD ₁ Similarly to the case, control point sequences for generating a cubic Bezier curve are respectively determined.

The A pattern feature quantity extraction unit 13, thus to white lumps WD determined by _(WD 1 _{~WD n)} and black lumps BD the _(BD 1 _{~BD n)} each control point sequence, a second A pattern feature quantity MP2 Data (hereinafter referred to as second pattern feature data) D15 (FIG. 7), and this is sent to the two-dimensional code conversion unit 14.

In this way, the pattern feature quantity extraction unit 13 extracts each white dama WD (WD _{1 to} WD _n ) and each black dama BD in order to accurately and rigorously extract the feature quantities of the white dama WD and the black dama BD. For each of (BD _{1 to} BD _n ), a method (hereinafter referred to as a second method) for determining a sequence of points (hereinafter referred to as control points) for generating a Bezier curve (bezier curve) on the basis of a dama outer periphery point. The second pattern feature amount MP2 is extracted based on the above-described extraction method.

On the other hand, A pattern feature quantity extracting unit 13, if it is the first verification mode, Kakushiro based on coded A pattern image MGX lumps WD _(WD 1 ~WD _n) and the black lumps BD _(BD 1 ~BD _n ) To calculate data of the first matching pattern feature value MPC1 (hereinafter referred to as the first matching feature value) by performing the same processing as in the first extraction method described above. It is generated as D18 (referred to as “print pattern feature data”) (FIG. 7) and is sent to the collation unit 15.

(3-1-5) Two-dimensional code conversion process The two-dimensional code conversion unit 14 stores the first pattern feature amount MP1 as the first pattern code BC1 (FIG. 2) on the original photographic paper OP, and the second The pattern feature amount MP2 is registered in the pattern code storage unit 6 in a state associated with the first pattern code BC1 as the second pattern code BC2 (FIG. 2).

  Specifically, the two-dimensional code conversion unit 14 rounds off the decimal point of the pattern feature amount (rectangular information value of each white dama WD and each black dama BD) of the supplied first pattern pattern data D16, and obtains the result. The two-dimensional barcode conversion processing based on the code character string information stored in the memory is applied to the pattern feature amount thus generated, thereby generating first pattern code data D2, and this is generated at a predetermined timing by the printer unit 5 To send.

  As a result, the first pattern code data D2 is printed as a first pattern code BC1 (FIG. 2) at a predetermined position of the photographic paper (original photographic paper OP) set on the photographic paper board in the printer unit 5. The original photographic paper OP (FIG. 2) is recorded.

  Further, the two-dimensional code conversion unit 14 rounds off the decimal point of the control point sequence (the control point sequence of each white dama WD and each black dama BD) of the supplied second pattern pattern data D17, and the result is obtained. The second pattern code data D3 is generated by performing two-dimensional bar code conversion processing based on the code character string information stored in the memory, and this data is associated with the first pattern code data D2. In this state, the data is sent to the pattern code storage unit 6 at a predetermined timing.

  As a result, the first pattern code data D2 and the second pattern code data D3 are associated with predetermined positions in the pattern code storage unit 6 as the first pattern code BC1 and the second pattern code BC2. It will be registered.

(3-1-6) Collation Processing In the first verification mode, the collation unit 15 uses the first collation pattern feature amount MPC1 extracted from the code-added photographic paper XPc and the original as shown in FIG. The first restored pattern feature value MPR1 restored based on the first pattern code BC1 stored in the photographic paper OP is collated.

  In practice, the collation unit 15 uses the white dama WD represented by the first collation pattern feature amount MPC1 (rectangular information value) in the first collation pattern feature amount data D18 supplied from the pattern feature amount extraction unit 13. And a black dama BD (hereinafter referred to as a collation dama) are represented by a first restored pattern feature quantity MGR1 (rectangular information value) based on the first pattern code data D2 provided from the scanner unit 4. A white dama WD and a black dama BD (hereinafter referred to as a restoration dama) are sequentially checked.

  Here, specific collation processing by the collation unit 15 will be described with reference to FIG. 14, but for convenience of explanation, here, collation processing of one collation dama and a restoration dama will be described.

FIG. 14 shows the positional relationship of rectangles represented by rectangle information values (center coordinates (x _c , y _c ), long side l, short side w, and angle θ between long side l and the horizontal axis). R _r is the rectangle (broken line) of the restored dama, S _r is the area represented by the long side l and the short side w of the restored dama, and g _r is the center coordinate (x _c , y _c ) of the restored dama The center represented, R is the rectangle (solid line) of the collation dama, S is the area represented by the long side l and short side w of the collation dama, and g is the center coordinate (x _c , y _c ) of the collation dama. Each center is shown.

  D is the following formula:

Indicates the distance between the centers g _r and g of the restoration dama and the matching dama calculated according to (hereinafter referred to as the center-to-center distance), and θ ′ is the angle between the long side l and the horizontal axis in the restoration dama The difference between θ and the angle θ formed between the long side l in the collation dama and the horizontal axis, that is, the difference in inclination between the rectangle _Rr and the rectangle R (hereinafter referred to as the difference between the rectangles) is shown in the figure. The ellipse indicates a collation dama.

In FIG. 14, the collation unit 15 determines that the restoration grinder center g _r exists in the collation dama rectangle R based on the rectangular information values of both the restoration dama and the collation dama, and the collation dama center g Is present in the rectangle R _{r of the} restoration dama.

Then, when both the centers g _r and g are present in the rectangles R and R _r , the collation unit 15 collates with the distance d between the centers, the inter-rectangular slope difference θ ′, and the area S _{r of the} restored dama. It is sequentially determined whether or not the difference from the dama area S (hereinafter referred to as dama area difference) is less than or equal to a predetermined threshold value.

  Here, the collation unit 15 determines that the restoration dama and the collation dama are the same dama if both are less than or equal to the threshold value, and if any one of them is greater than or equal to the threshold value Therefore, it is determined that the restoration dama and the collation dama are not the same dama.

However, as shown in FIG. 15 in which the same reference numerals are given to corresponding parts to FIG. 14, when the rectangles R _r and R of both the restoration dama and the collation dama are both close to a square, the inter-rectangular slope difference θ ′ is approximately 90. Therefore, there may occur a situation in which it is determined that they are different from each other even though they are the same.

Accordingly, as a countermeasure for preventing such a misjudgment, the collation unit 15 has both the ratio of the long side l _r and the short side w _r of the restoration dama and the ratio of the long side l and the short side w of the collation dama “ If the slope difference between the rectangles θ _r −θ (that is, θ ′ in FIG. 14) is equal to or greater than the threshold value and the dama area difference is equal to or less than the threshold value, Are determined to be the same.

  In this way, the collation unit 15 uses each collation dama represented by the first collation pattern feature amount MPC1 (rectangular information value) extracted from the code-added photographic paper XPc (FIG. 2B) and the original. Each restored dama represented by the first pattern feature quantity (rectangular information value) based on the first pattern code BC1 stored on the photographic paper OP is collated.

  Then, the collation unit 15 determines that the code-added photographic paper XPc corresponding to the code-pattern image MGX is a valid original photographic paper OP only when a match rate higher than a predetermined match rate is obtained as a result of this check. For example, the fact that the code-added photographic paper XPc is the original photographic paper OP, that is, that it is guaranteed to be an original is notified via a predetermined display unit (not shown) or the like.

  In the second verification mode, the collation unit 15 stores the second collation pattern feature amount MPC2 extracted from the code-added photographic paper XPc and the original photographic paper OP as shown in FIG. The second restored pattern feature value MPR2 restored based on the second pattern code BC2 associated with the first pattern code BC1 is collated.

  Specifically, the collation unit 15 detects the second pattern code BC2 associated with the first pattern code BC1 by searching the supplied pattern code storage unit 6 in the pattern code storage unit 6. Then, the collation unit 15 is caused to read the second pattern code data D19.

  The collation unit 15 restores the second restored pattern feature value MPR2 by performing an inverse two-dimensional code conversion process on the second pattern code BC2 of the second pattern code data D3. Based on the control point sequence of the restored pattern feature value MPR2, the white dama corresponding to the original white dama WD (hereinafter referred to as a restored white dama) and the black dama corresponding to the original black dama BD (hereinafter referred to as this). Is called a restoration black dama).

Here, the reconstruction method of the restored white dama and the restored black dama will be specifically described. For convenience of explanation, here, a case where the restored white dama corresponding to the white dama WD ₁ described in FIG. 13 is restored will be described. To do.

As shown in FIG. 16, the collation unit 15 controls the control point sequence P ₁ -C ₂ -C for the white dama WD ₁ (the hatched area surrounded by the broken line in FIG. 16) extracted by the pattern feature amount extraction unit 13. ₃ -P _4, the _{P 4 -C 5 -C 6 -P 7} , P 7 -C 8 -C 9 -P 10, P 10 -C 11 -C 12, the following equation

Respectively, corresponding Bezier curves Bc1, Bc2, Bc3, and Bc4 are generated. Note that “CP” in the equation (6) is a control point, and “B ⁿ _i (t)” is called a Bernstein function.

Is defined by

  Then, as shown in FIG. 17, the collation unit 15 generates a restored white drum by filling the area surrounded by the Bezier curves Bc1, Bc2, Bc3, and Bc4 with a predetermined single luminance value.

In this way, the matching unit 15 is adapted to generate a restoring white lumps corresponding to white lumps WD _1.

The collating unit 16, as in the case of white lumps WD _1, based on the control points of the white lumps _WD 2 _{~WD n} in the second restoration A pattern feature quantity mgr2, the white lumps _WD 2 _{~WD n} And a restored black dama corresponding to the black dama BD based on the control point sequence for the black dama BD (BD _{1 to} BD _n ) in the second restored pattern feature quantity MGR2. Has been made to generate.

Then, the collation unit 15 and the white damascene in the second collation feature amount MPC2 of the second collation pattern feature amount data D16 supplied from the area division unit 13 at this time are generated. Black in the second matching feature quantity MPC2 of the second matching pattern feature quantity data D16 supplied from the area dividing unit 13 between the WD (WD _{1 to} WD _n ) and the restored black dama. The phase-only correlation value C _poc between the dama BD (BD _{1 to} BD _n ), R (x, y) for the restored white dama and restored black dama pixels, and the white dama WD and black dama BD pixels If D (x, y), the two-dimensional Fourier transform is F, and the two-dimensional inverse Fourier transform is F- ¹ , then

Calculate according to

Here, when the phase-only correlation value C _poc less than or equal to a predetermined threshold value is obtained, the collation unit 15 uses the code-added photographic paper XPc (FIG. 4) placed on the placement table of the scanner unit 4 at this time. For example, it is determined that the code-added photographic paper XPc is the original photographic paper OP, that is, a denial that it is an original is notified through a predetermined display unit (not shown) or the like. It is made like that.

On the other hand, when a phase-only correlation value C _poc higher than a predetermined threshold is obtained, the collation unit 15 at this time places the code-added photographic paper XPc (FIG. 4) placed on the placement table of the scanner unit 4. ) Is a legitimate original photographic paper OP, for example, a predetermined display unit (not shown) or the like is provided to guarantee that the code-added photographic paper XPc is the original photographic paper OP, that is, the original. To be notified through.

(3-2) Processing in the control unit in the second embodiment Here, in the code print registration mode, the first verification mode, and the second verification mode in the control unit 2 in the second embodiment. When the processing contents are functionally classified, as shown in FIG. 18, a low-frequency component extraction unit 20, an image separation unit 21, a region division unit 22, a pattern feature amount extraction unit 23, and a two-dimensional code conversion It can be divided into a unit 24 and a collation unit 25.

  In this case, in the code print registration mode, the control unit 2 applies the low frequency component extraction unit 20, the image separation unit 21, the region division unit 22, the pattern feature amount to the pattern image data D1 given from the scanner unit 4. Various processes are sequentially performed through the extraction unit 23 and the two-dimensional code conversion unit 24, and the first pattern code data D2 obtained as a result is sent to the printer unit 5, and the first pattern code data D2 and the second pattern code data D2 The pattern code data D3 is sent to the pattern code storage unit 6 in a state of being associated therewith.

  In the first verification mode, the control unit 2 applies the low frequency component extraction unit 20, the image separation unit 21, the region division unit 22, and the pattern feature to the code-attached pattern image data D4 given from the scanner unit 4. After performing various processes sequentially through the quantity extraction unit 23, the verification unit 15 performs a verification process based on the processing result and the first pattern code data D2 given from the scanner unit 4.

  Further, in the second verification mode, the control unit 2 applies the low frequency component extraction unit 20, the image separation unit 21, the region division unit 22, and the pattern feature to the code-attached pattern image data D4 given from the scanner unit 4. After performing various processes sequentially through the quantity extraction unit 23, the verification unit 15 performs a verification process based on the processing result and the second pattern code data D3 given from the pattern code storage unit 6. Yes.

  Hereinafter, low frequency component extraction processing by the low frequency component extraction unit 20, image separation processing by the image separation unit 21, image division processing by the region division unit 22, pattern pattern extraction processing by the pattern pattern extraction unit 23, two-dimensional The two-dimensional code conversion process by the code conversion unit 24 and the verification process by the verification unit 25 will be described in detail.

(3-2-1) Low Frequency Component Extraction Processing In the code print registration mode, the low frequency component extraction unit 20, as shown in FIG. 19, for example, a pattern image MG (FIG. 2) on the original photographic paper OP (FIG. 2). 19 (A)), a low-frequency component image (hereinafter referred to as a first low-frequency pattern image) MGL1 (FIG. 19B) is extracted.

  Specifically, the low frequency component extraction unit 20 performs a Fourier transform process on the pattern image MG in the pattern image data D1 provided from the scanner unit 4 based on the region position information stored in the internal memory. Generate component data.

  Subsequently, the low frequency component extraction unit 20 performs the inverse Fourier transform process on the data of the frequency component after setting the data value of the high frequency component equal to or higher than a predetermined threshold to “0”, thereby the first low frequency pattern. Data of the image MGL1 (hereinafter referred to as first low-band pattern image data) D30 (FIG. 18) is generated.

  Then, the low frequency component extraction unit 20 sends the first low frequency pattern image data D30 generated at this time to the pattern feature value extraction unit 21.

  Further, in the code print registration mode, the low-frequency component extraction unit 20 uses the pattern image MG (FIG. 19A) to display the feature of the pattern image in more detail than the first low-frequency pattern image MGL1 described above. A low-frequency component pattern image (hereinafter referred to as a second low-frequency pattern image) MGL2 (FIG. 19C) is extracted.

  Specifically, the low frequency component extraction unit 20 performs a Fourier transform process on the pattern image MG in the pattern image data D1 provided from the scanner unit 4 based on the region position information stored in the internal memory. Generate component data.

  Subsequently, the low frequency component extraction unit 20 sets “0” to a data value equal to or higher than the threshold value of the high frequency component compared to the first low frequency pattern image MGL1 described above for the frequency component data. By performing an inverse Fourier transform process, data (hereinafter referred to as second low-frequency pattern image data) D31 (FIG. 18) of the second low-frequency pattern image MGL2 is generated.

  Then, the low frequency component extraction unit 20 sends the second low frequency pattern image data D31 generated at this time to the pattern feature value extraction unit 21.

  On the other hand, in the first verification mode, the low-frequency component extraction unit 20 uses the code image with a low-frequency component (FIG. 19A) on the code-added photographic paper XPc (FIG. 3). Hereinafter, MGXL1 (FIG. 19B) is extracted, which is referred to as a first code-added low-frequency pattern image.

  Specifically, the low-frequency component extraction unit 20 applies the above-mentioned code pattern image MGX in the code pattern image data D4 provided from the scanner unit 4 based on the region position information stored in the internal memory. The data of the coded low-frequency pattern image MGXL1 extracted by performing the same Fourier transform process and inverse Fourier transform process as in the case of the first low-frequency pattern image MGL1 (hereinafter referred to as the first code-added low-frequency pattern image D32) is generated.

  Then, the low frequency component extraction unit 20 sends the first code-added low frequency pattern image D32 generated at this time to the pattern image synthesis unit 21.

  On the other hand, in the second verification mode, the low-frequency component extraction unit 20 uses the code-added pattern image MGX (FIG. 19A) in more detail than the first code-added low-pass pattern image MGXL1. A pattern image of a low frequency component representing the characteristics of the pattern image (hereinafter referred to as second code-added low frequency pattern image data) MGXL2 (FIG. 19C) is extracted.

  Specifically, the low-frequency component extraction unit 20 applies the above-mentioned code pattern image MGX in the code pattern image data D4 provided from the scanner unit 4 based on the region position information stored in the internal memory. The data of the coded low-frequency pattern image MGXL2 extracted by performing the same Fourier transform process and inverse Fourier transform process as in the case of the second low-frequency pattern image MGL2 (hereinafter, this is referred to as the second coded low-frequency pattern image). D33) is generated.

  The low frequency component extraction unit 20 is configured to send the second low frequency pattern image D33 with code generated at this time to the pattern image synthesis unit 21.

(3-2-2) Image Separation Processing The image separation unit 21 includes the first low-frequency pattern image MGL1, the first low-frequency pattern image MGL2, and the coded low-frequency pattern extracted by the low-frequency component extraction unit 10. The image MGXL2 or the coded low band pattern image MGXL2 is separated into the white component pattern image MGW and the black component pattern image MGB by performing the same process as the image separation process in the first embodiment described above. Then, it is sent to the area dividing unit 22 as white component pattern image data D34 and black component pattern image data D35.

(3-2-3) Area division processing The area division unit 22 performs a process similar to the area division process in the first embodiment described above, thereby performing a plurality of patterns on the white component pattern image MGW. Are divided into white dama WD (WD _{1 to} WD _n ), and the black pattern image MGB is divided into a plurality of black dama BDs (BD _{1 to} BD _n ), which are divided into white dama data D36 and black dama data. The result is sent to the pattern feature quantity extraction unit 23 as D37.

(3-2-4) Pattern Feature Value Extraction Processing The pattern feature value extraction unit 23, in the code print registration mode, each white dama WD (WD _{1 to} WD _n ) and each black based on the first low-frequency pattern image MGL1. The feature values of the shape in the dama BD (BD _{1 to} BD _n ) are calculated by performing the same processing as in the first extraction method in the first embodiment described above, so that the first pattern feature is obtained. Quantity data D38 (FIG. 18) is generated and sent to the two-dimensional code conversion unit 24.

Also, A pattern feature quantity extraction unit 23, the code printing registration mode, in Kakushiro lumps WD based on the second low-pass A pattern image MGL2 _(WD 1 ~WD _n) and the black lumps BD _(BD 1 ~BD _n) The second pattern feature quantity data D39 (FIG. 18) is generated by calculating the feature quantity of the shape by performing the same processing as the first extraction method in the first embodiment described above. This is sent to the two-dimensional code converter 24.

On the other hand, in the case of the first verification mode, the pattern feature amount extraction unit 23 sets each white dama WD (WD _{1 to} WD _n ) and each black dama BD (based on the first code-added low-band pattern image MGXL1). BD _{1 to} BD _n ) are calculated by performing the same processing as the first extraction method in the first embodiment described above, so that the first matching pattern feature amount is calculated. Data) D40 (FIG. 18) is generated and sent to the collation unit 25.

On the other hand, in the case of the second verification mode, the pattern feature quantity extraction unit 23 uses each white dama WD (WD _{1 to} WD _n ) and each black dama BD (based on the second low-frequency pattern image MGXL2 with code). BD _{1 to} BD _n ) are calculated by performing the same processing as the first extraction method in the first embodiment described above, so that the second matching pattern feature amount is calculated. Data) D41 (FIG. 18) is generated and sent to the collation unit 25.

(3-2-5) Two-dimensional code conversion process The two-dimensional code conversion unit 24 performs the same process as the two-dimensional code conversion process in the first embodiment described above, thereby the first pattern feature amount MP1. Is stored in the original photographic paper OP as the first pattern code BC1 (FIG. 2), and the second pattern feature amount MP2 is associated with the first pattern code BC1 as the second pattern code BC2 (FIG. 2). In this state, it is registered in the pattern code storage unit 6.

(3-2-6) Collation Processing The collation unit 25 selects the first extracted from the code-added photographic paper XPc (FIG. 2B) when simplicity or speed is given priority. The first restored pattern feature value MPR1 (based on each matching dama represented by the matching pattern feature value MPC1 (rectangular information value) and the first stored pattern code BC1 of the original photographic paper OP ( The code-added photographic paper XPc corresponding to the code-attached pattern image MGX is obtained only when the restored dama represented by the (rectangular information value) is collated and a matching rate higher than a predetermined matching rate is obtained as a matching result. It is determined that the original photographic paper OP is a legitimate original photographic paper OP, and the code-added photographic paper XPc is guaranteed to be the original.

  Further, the collation unit 25 uses the first collation pattern feature value MPC1 (rectangular information value) extracted from the code-added photographic paper XPc (FIG. 3) when notarization is given first priority. Each matching dama represented, and each restored dama represented by a first restored pattern feature value MPR1 (rectangular information value) based on the first stored pattern code BC1 of the original photographic paper OP. Only when a matching rate higher than a predetermined matching rate is obtained as a matching result, it is determined that the code-added photographic paper XPc corresponding to the code-attached pattern image MGX is a valid original photographic paper OP, and the code It is ensured that the attached photographic paper XPc is the original.

(4) Usage Example of Original Guarantee Device in the Present Embodiment Next, a usage example of the original guarantee device 1 in the present embodiment will be described.

  In this case, the original guarantee device 1 uses a check as the photographic paper on which the print content is printed. Then, in the original guarantee device 1, as shown in FIG. 20, among the original guarantee devices 1, a device (hereinafter referred to as “the first verification mode” and hardware that is necessary for executing the first verification mode). This is referred to as a first device), a code printing registration mode, a second verification mode, and a device (hereinafter referred to as the first device) equipped with the hardware necessary for executing these modes and the pattern code storage unit 6. This is a more desirable form when applied separately.

  In practice, the second device EQ2 is installed in a bank or the like, for example, and extracts the first pattern feature amount MP1 from the pattern image MG in the specified area AR of the check CH, A second pattern feature amount MP2 representing a detailed feature amount compared to the first pattern feature amount MP1 is extracted, and the first pattern feature amount MP1 is printed on the check CH as a first pattern code BC1. The second pattern feature amount MP2 is registered as the second pattern code BC2 in the pattern code storage unit 6 in association with the first pattern code BC1.

  The first apparatus EQ1 is installed in, for example, a company or a home. For example, when the check CH is drawn out from the check CH, or when the check CH is distributed thereafter. That is, when simplicity or speed is given first priority, the code-added photographic paper is checked by comparing the first matching pattern feature value MPC1 with the first restored pattern feature value MPR1. It is possible to easily and quickly verify that XPc is different from the original.

  Further, the second apparatus EQ2, for example, when it is exchanged for cash based on the drawn check CH, that is, when notarity is given first priority, the first pattern feature amount MP1. In order to extract the second pattern feature quantity MP2 representing a detailed feature quantity compared to the second, the second pattern feature quantity MPC2 is collated with the second restored pattern feature quantity MPR2, so that the second The matching pattern feature quantity MPC2 and the second restored pattern feature quantity MPR2 can be verified precisely and strictly.

  Note that the original guarantee device 1 can be used as it is without being separated into the first device EQ1 and the second device EQ2.

  As shown in FIG. 21, if the pattern code storage unit 6 is separated from the second device EQ2 and used as the server device EQ3, a plurality of first devices EQ1 and a plurality of second devices EQ2 are used. Further, a large-scale system can be constructed as shown in FIG. 20, which includes the server apparatus EQ3.

(5) Operation and effect according to this embodiment In the above configuration, the original guarantee device 1 extracts the first pattern feature amount MP1 from the pattern image MG of the pattern in the designated area AR of the original photographic paper OP. At the same time, a second pattern feature quantity MP2 representing a detailed feature quantity compared to the first pattern feature quantity MP1 is extracted, and the original photographic paper OP is used with the first pattern feature quantity MP1 as the first pattern code BC1. On the other hand, the second pattern feature amount MP2 is registered as the second pattern code BC2 in association with the first pattern code BC1.

  Accordingly, when simplicity or speed is given first priority, the code-added photographic paper is checked by comparing the first matching pattern feature value MPC1 with the first restored pattern feature value MPR1. When the XPc is different from the original, it can be verified easily and quickly, and when notarity is the first priority, a detailed feature amount is compared with the first pattern feature amount MP1. In order to extract the second pattern feature quantity MP2 to be represented, the second matching pattern feature quantity MPC2 is compared with the second matching pattern feature quantity MPC2, and the second restored pattern feature quantity MPR2 is compared. The second restored pattern feature value MPR2 can be verified precisely and strictly.

  According to the above configuration, the first pattern feature amount MP1 is extracted from the pattern image MG of the pattern in the designated area AR of the original photographic paper OP, and more detailed than the first pattern feature amount MP1. A second pattern feature value MP2 representing the feature value is extracted, and the first pattern feature value MP1 is printed as the first pattern code BC1 on the original photographic paper OP, while the second pattern feature value MP2 is set to the second pattern feature value MP2. In the case where simplicity or quickness is given first priority by registering the pattern code BC2 in association with the first pattern code BC1, the first matching pattern feature amount MPC1 and the first pattern code MP1 By comparing with the restored pattern feature value MPR1 of 1, it is possible to easily and quickly verify that the code-added photographic paper XPc is different from the original, and when notarity is given priority first. The second In order to extract the second pattern feature quantity MP2 representing a detailed feature quantity compared to the pattern feature quantity MP1, the second pattern feature quantity MPC2 for collation and the second restored pattern feature quantity MPR2 are collated. As a result, the second matching pattern feature value MPC2 and the second restored pattern feature value MPR2 can be verified precisely and strictly, thus appropriately protecting the print contents printed on the original photographic paper OP. be able to.

(6) Other Embodiments In the above-described embodiments, a plurality of papers with predetermined contents are used as objects for printing various contents such as documents and photographs, and the contents of the prints are printed. Although the case where photographic paper is applied has been described, the present invention is not limited to this, and may be, for example, a work of art such as a picture, or a personal signature such as a signature. Any paper with contents attached thereto is acceptable.

  In the above-described embodiment, as the first extraction method, the white dama WD and the black dama BD are approximated to a rectangle in order to easily and quickly extract the feature amounts of the white dama WD and the black dama BD. However, the present invention is not limited to this, and the extraction method for extracting the features of the pattern image may be changed in accordance with the contents attached to the paper.

  In this case, as an example of use of the original guarantee device 1 in the present embodiment, the case where it is applied to a check CH as a paper with a predetermined content has been described, but the importance is low as a paper with a predetermined content. In the case of a document that seems to be, for example, the first extraction method is a simple extraction method compared to the extraction method based on the above-described rectangular approximation, for example, a low-frequency pattern image is used as the first extraction method. In the case of a document that seems to be highly important as paper with a predetermined content, for example, compared to the extraction method based on the above rectangle approximation such as the extraction method based on the circle approximation or the ellipse approximation. The detailed extraction method may be used as the first extraction method.

  In this way, verification is performed while appropriately adjusting the waiting time of the user according to the importance of the content by changing the extraction method for extracting the feature of the pattern image according to the importance of the content attached to the paper. Thus, the contents attached to the paper can be appropriately protected.

  Furthermore, in the above-described embodiment, low-frequency component processing, image separation processing, region division processing, and pattern feature amount extraction processing are sequentially performed on the pattern image captured by the scanner unit 4 (imaging unit). However, the present invention is not limited to this, and it is not always necessary to take an image, or it is not necessary to perform all of these processes, and processes other than these processes are performed. The feature may be added or changed. In short, the feature may be acquired based on a signal obtained from the pattern projection light obtained by transmitting through the paper.

  In this case, for example, the region division processing and the pattern feature amount extraction processing are executed without executing the low frequency component processing or the image separation processing, and conversely, the low frequency component processing and the image separation processing are executed. Only the white component pattern image WIM (FIG. 9B) and the black component pattern image BIM (FIG. 9C) obtained as a result are extracted as pattern feature amounts, or Only the frequency component processing may be executed, and the low-frequency pattern image MGL (FIG. 9A) obtained as a result may be extracted as the pattern feature amount.

  Further, the pattern image captured by the scanner unit 4 (imaging means) is divided into, for example, 5 × 5 vertical and horizontal image areas, and one of the extracted areas is extracted as a pattern feature amount. Alternatively, the feature quantity may be extracted after dividing the pattern in one of the extracted areas in the same manner as the area dividing process in the area dividing unit 13.

  Further, as the image separation processing, the white component pattern image WIM and the black component pattern image BIM are separated, but the white component pattern image WIM (FIG. 9B) or the black component pattern image BIM (FIG. 9C Only one of)) may be extracted.

  Furthermore, as the pattern feature quantity extraction process, each of the white dama WD and the black dama BD is approximated to a rectangle, but instead, it may be approximated to various other shapes such as a spherical shape.

  Further, the pattern image is acquired from the designated area AR (FIG. 2) of the photographic paper. However, the present invention is not limited to this, and the pattern image may be obtained from a plurality of designated areas on the photographic paper. You may make it acquire from the whole paper.

  Furthermore, in the above-described embodiment, the case has been described in which the pattern feature amount is printed on the photographic paper (original photographic paper OP) as the pattern code BC1 (FIG. 2). However, the present invention is not limited to this. For example, holes or braille corresponding to the pattern feature amount may be provided on the photographic paper, or the pattern feature amount may be directly described on the photographic paper.

In this case, instead of the pattern pattern (pattern feature value), the white component pattern image WIM (FIG. 9B), the black component pattern image BIM (FIG. 9C), and the white dama WD (WD _{1 to} WD _n). ), Black dama BD (BD _{1 to} BD _n ) or a combination thereof may be directly printed on the photographic paper. In short, various other pattern information acquired by the extracting means described above is stored. Can do.

  Further, in the above-described embodiment, the case has been described in which the pattern feature amount is verified by collating with the first extraction method or the second extraction method described above, but the present invention is not limited to this, and other than this. You may make it collate by the method of. In this case, the same effect as that of the above-described embodiment can be obtained.

  Further, in the above-described embodiment, as a dividing means for dividing the pattern on the photographic paper into a predetermined unit area, for example, from the pattern image obtained as a result of reading by the scanner unit 4, the low-frequency component pattern image MGL (FIG. 8 (B)) is extracted, and the low-frequency component pattern image MGL is separated into a white component pattern image WIM (FIG. 9B) and a black component pattern image BIM (FIG. 9C). The pattern included in the image WIM and the black component pattern image BIM is divided into white dama WD and black dama BD by sequentially connecting pixels of 8 neighboring pixels. However, the present invention is not limited to this, and is not necessarily limited thereto. It is not necessary to extract the low-frequency component pattern image MGL, and it is not always necessary to separate it into the white component pattern image WIM and the black component pattern image BIM. Further, white pixels (black pixels) of neighboring pixels are sequentially connected. The Thus, it is not necessary to divide the white dama WD (black dama BD). In short, the pattern included in the pattern image obtained as a result of reading by the scanner unit 4 is divided into predetermined unit areas. Various techniques can be applied.

  Furthermore, in the above-described embodiment, for each divided area, a plurality of points for generating a curve that approximates the outline is determined based on the points on the outline of the area, and these points are used as pattern information. As the extraction means for extracting as follows, the lattice size and the order of the Bezier curve are switched according to the total area of the white dama WD and black dama BD, and the outer circumference of the white dama WD and black dama BD intersecting with the square lattice of the switched lattice size. Although the case where the control point sequence for generating the Bezier curve based on the upper point is extracted as the pattern information has been described, the present invention is not limited to this, and the largest of the white dama WD and the black dama BD. Switching may be performed according to the dama area, and the lattice size and / or the order of the Bezier curve may be set as fixed values.

  Further, the points intersecting the square lattice are the points on the outer periphery of the white dama WD and the black dama BD. However, the present invention is not limited to this. For example, a certain reference point on the outer periphery of the white dama WD and the black dama BD is determined. A point that intersects with the circle centered on the reference point is a point on the outer periphery of the white dama WD and black dama BD, and a point that intersects the circle centered on this point is the outer periphery of the next white dama WD and black dama BD. You may make it go to the upper point. The diameter or radius of the circle at this time may be switched according to the dama total area of the white dama WD and the black dama BD.

  The control point sequence for generating the Bezier curve is extracted as the pattern information. However, the present invention is not limited to this, and various other curves such as a rational Bezier curve, a B-spline curve, or a rational B-spline curve are used. A control point sequence for generation may be extracted as pattern information. In this case, the same effect as that of the above-described embodiment can be obtained.

  Further, in the above-described embodiment, the case where the configuration shown in FIG. 6 is applied as the original prevention device has been described. However, the present invention is not limited to this, and various other configurations can be applied. .

  In this case, a program that causes the control unit to execute various processes shown in FIG. 7 or a part of the processes is installed in an existing apparatus that handles paper such as a copier or a newly manufactured apparatus. May be.

  Further, a program that causes the control unit to execute various processes shown in FIG. 7 or a part of the processes may be applied to an apparatus having a hardware configuration.

  The present invention can be used when paper is used as various media such as commodity exchange media such as money, content certification media such as certificates, or information storage media such as personal works.

It is a basic diagram which shows the pattern of paper. It is a basic diagram with which it uses for description of the relevance guarantee method using a pattern feature-value. It is a basic diagram with which it uses for description of the relevance guarantee method using a pattern feature-value. It is a basic diagram with which it uses for description of the relevance guarantee method using a pattern feature-value. It is an approximate line figure used for explanation of verification of an original. It is a block diagram which shows the structure of the original guarantee apparatus by this Embodiment. It is a block diagram with which it uses for description of the process in the control part in 1st Embodiment. It is a basic diagram with which it uses for description of extraction of a low frequency component. It is an approximate line figure used for explanation of separation of an image. It is a basic diagram used for description of the division of a white dama (black dama). It is a basic diagram with which it uses for description of the removal of a small dama (white component pattern image). It is a basic diagram with which it uses for description of extraction of the 1st pattern feature-value. It is a basic diagram with which it uses for description of the determination of a control point example. It is an approximate line figure used for explanation of collation of a dama. It is an approximate line figure used for explanation of collation of a dama. It is a basic diagram with which it uses for description of the production | generation of a Bezier curve. It is a basic diagram with which it uses for description of the production | generation of a reconstruction dama. It is a block diagram with which it uses for description of the process in the control part in 2nd Embodiment. It is a basic diagram with which it uses for description of extraction of a low frequency component. It is a basic diagram with which it uses for description of the usage example of an original guarantee apparatus. It is a basic diagram with which it uses for description of the usage example of an original guarantee apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Original guarantee apparatus, 2 ... Control part, 4 ... Scanner part, 4a ... Code reader, 5 ... Printer part, 6 ... Pattern code storage part, 10, 20 ... Low frequency component extraction part , 11, 21... Image separation unit, 12, 22... Area division unit, 13, 23... Pattern feature amount extraction unit, 14, 24... Two-dimensional code conversion unit, 15, 25. D1: Pattern image data, D2: First pattern code data, D3: Second pattern code data, D4: Pattern image data with code, D5: Print content image data, D10: Low band pattern Image data, D11: Coded low-frequency pattern image data, D12 ... Coded pattern image data, D13 ... Pattern pattern data, D14 ... Pattern image data for verification, D15 ... Pattern code data, D16 ... Re Composition pattern image data, O …… Original photographic paper, XPc …… Coded photographic paper, AR …… Specified area, BC …… Pattern code, MG …… Pattern image, MGX …… Pattern image with code, MGL …… Low-band pattern image, MGXL… ... Low-band pattern image with code, MGR: Restoration pattern image, MGC: Pattern image for verification, MP: Pattern feature amount, MGW: White component pattern image, MGB: Black component pattern image, WD (WD ₁ ˜WD _n ) …… White Dama, BD (BD _{1 to} BD _n ) …… Black Dama.

Claims (6)

Imaging means for imaging a unique pattern on the paper with a predetermined content;
First extraction means for extracting features of the pattern image captured by the imaging means;
Second extraction means for extracting features from the pattern image in more detail than the first feature extracted by the first extraction means;
Storage means for storing the first feature on the paper;
An information generation apparatus comprising: a registration unit that registers the second feature extracted by the second extraction unit in a predetermined storage unit. The first extraction means includes
The information generation apparatus according to claim 1, wherein an extraction method for extracting features of the pattern image is changed according to the contents attached to the paper. Dividing means for dividing the pattern image picked up by the image pickup means into a predetermined unit area;
The first extraction means includes
For each of the areas divided by the dividing means, approximate the outline of the area to a rectangle, and extract the feature of the rectangle as the first feature,
The second extraction means includes
For each of the regions divided by the dividing means, a plurality of points for generating a curve that approximates the contour is determined based on the points on the contour of the region, and these points are determined as the second point. The information generation device according to claim 1, wherein the information generation device is extracted as a feature. The second extraction means includes
Extracting the second feature based on a second low-frequency pattern image composed of low-frequency components from the pattern image,
The first extraction means includes
The first feature is extracted from the pattern image based on a first low-frequency pattern image having a low-frequency component compared to the second low-frequency pattern image. The information generation device described in 1. A first step of imaging a unique pattern on a paper with a predetermined content;
A second step of extracting a feature of the pattern image captured in the first step as a first feature and extracting a feature more detailed than the first feature in the pattern image as a second feature;
And a third step of storing the first feature on the paper and registering the second feature in a predetermined storage means. A first process for capturing an image of a unique pattern on a paper with a predetermined content;
A second process for extracting a feature of the pattern image captured in the first step as a first feature and extracting a feature more detailed than the first feature in the pattern image as a second feature;
A program for storing the first feature on the paper and executing a third process for registering the second feature in a predetermined storage unit.