JP2008172758A - Image processing method, image processing device and image processing program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing method, an image processing device and an image processing program capable of efficiently performing information extraction when embedding information in an image by using plural kinds of methods. <P>SOLUTION: The image processing method for an image processing device to embed information in an image comprises: an information embedding step of embedding target information in the image by using one or more methods selected from plural information embedding methods; and a method identification information embedding step of embedding method identification information for identifying the selected one or more methods in the image. In the method identification information embedding step, the method identification information is embedded in a method allowing an amount of the embedded information to be less than an amount of the embedded information in each of the selected one or more methods. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an image processing method, an image processing apparatus, and an image processing program, and more particularly to an image processing method, an image processing apparatus, and an image processing program that execute embedding or extraction of information with respect to an image.

  In recent years, due to improvements in image processing technology and image forming technology, when copying banknotes and securities using a digital color copier, the copy is faithful enough that the original copy cannot be easily distinguished from the original copy. It has become possible. For this reason, it is necessary to take measures so that special manuscripts such as banknotes and securities cannot be copied at all or cannot be copied correctly.

  Further, for example, even in the case of general documents other than special manuscripts such as banknotes and securities, there are many confidential documents that are prohibited from being output such as copying from the viewpoint of maintaining confidentiality of the document contents. It is necessary to take measures to prevent such confidential documents from being copied at all, or from being copied correctly.

  For this reason, various inventions have been made in the past that aim to exert restrictions on copying special manuscripts and confidential documents. For example, Patent Literature 1 proposes a technique for prohibiting output when a predetermined dot pattern is detected from an image read by a scanner. Therefore, by embedding a predetermined dot pattern in a confidential document or the like that prohibits copying, the reproduction process can be effectively prevented.

  However, according to the invention described in Patent Document 1, it is possible to effectively prevent copying of a confidential document, but the amount of information that can be embedded in the document is whether it is a confidential document, that is, 1 bit. Minutes. However, in order to realize a highly flexible information security function such as switching the permission / prohibition of copying according to the user's job, for example, in combination with user authentication, an information amount of at least several bits is required. In addition to controlling output prohibition at the time of copying, embedding tracking information that can identify the outflow source when a copy that does not have the output prohibition data detection function such as the dot pattern described above flows out For this purpose, an information amount of about 100 bits or more is required.

  Therefore, the present applicant has already proposed an information embedding method using a background dot pattern that enables extraction of information of about 100 bits (for example, Patent Document 2). There are various other information embedding methods that can extract information of about 100 bits, and each has advantages and disadvantages. For example, the method using the background dot pattern repeatedly embeds the dot pattern in the background portion of the document, so the readability of the document is reduced to some extent, but it is strong against concealment. On the other hand, a method of embedding a code image in a specific area, such as a general bar code or a two-dimensional bar code, is rich in versatility, but has a problem that it is easily hidden during copying in a copy prohibition control application. Further, in the case of a document image, according to a method for changing the character spacing (for example, Non-Patent Document 1) and a method for deforming the character outline (for example, Non-Patent Document 2), it is hardly seen by humans. Information can be embedded.

Therefore, if one or a plurality of embedding methods can be selected from various methods according to the purpose of embedding information, improvement in convenience is expected.
JP 2004-274092 A JP 2006-287902 A Tomio Amano and Yuki Hirayama, “Digital watermark embedding method in page description using layout structure” (September 17, 1998), published by Information Processing Society of Japan, pages 45-50 Shuichi Sugoi, Masayoshi Kamijo, “Digital watermarking method for lettering using character shapes” (November 25, 1999), published by The Institute of Electronics, Information and Communication Engineers, pp. 2175-2177

  However, if the multi-bit information embedding method can be arbitrarily selected from a plurality of embedding methods, it is not possible to specify which method is embedded when extracting information. Therefore, it is necessary to perform information extraction processing for all possible embedding methods, and there is a problem that the performance of the information extraction processing deteriorates.

  In addition, information extraction processing is also executed for general documents in which a large number of bit information is not embedded in the first place, and there is a problem that the performance deteriorates to copy processing of general documents that do not require copy control or the like.

  In general, the larger the amount of information embedded in a document of a certain area, the larger the memory and processing amount required for the extraction process. Therefore, it is possible to extract information embedded by the above-described multi-bit information embedding method in real time at the time of copying with a copying machine (that is, to perform information extraction processing in parallel while scanning an image line by line with a line sensor). It is very difficult. Therefore, in order to extract such information at the time of copying, it is necessary to stop the output processing until the entire image or image information necessary for information extraction is captured in the frame memory. Degradation of the material can become so great that it cannot be ignored.

  The present invention has been made in view of the above points, and in the case where information is embedded in an image by a plurality of types of methods, an image processing method and image processing capable of efficiently performing the information extraction processing An object is to provide an apparatus and an image processing program.

  Accordingly, in order to solve the above-described problem, the present invention provides an image processing method in which an image processing apparatus embeds information in an image, and one or more selected from a plurality of methods for embedding information in the image An information embedding procedure for embedding the information by using the method, and a method identification information embedding procedure for embedding method identification information for identifying the one or more selected methods in the image. In the identification information embedding procedure, the method identification information is embedded by a method that embeds less information than any method used in the information embedding procedure.

  In such an image processing method, when information is embedded in an image by a plurality of types of methods, the information extraction process can be performed efficiently.

  In order to solve the above problem, the present invention may be an image processing apparatus that executes the image processing method or an image processing program that causes a computer to execute the image processing method.

  According to the present invention, it is possible to provide an image processing method, an image processing apparatus, and an image processing program capable of efficiently performing information extraction processing when information is embedded in an image by a plurality of types of methods. it can.

  First, an outline of the present embodiment will be described. In this embodiment, a case will be described in which information may be embedded in an image by three types of methods. Here, “image” can be recognized by human vision regardless of the state, such as what is recorded (printed) on a medium such as paper, or electromagnetically recorded as electronic data. Information, or general data that represents the information.

In the present embodiment, the state of an image is roughly classified into the following four states.
(1) A state in which information is embedded by any one of three methods.
(2) A state in which information is embedded by any two of the three methods.
(3) A state in which information is embedded by all three methods.
(4) A state in which information is not embedded by any method.

  Therefore, in the present embodiment, information for identifying by which method the information is embedded (that is, information for identifying the states (1) to (4) above. Hereinafter, “embedding method identification”. Information ”) is embedded in the image. When extracting information embedded by any of the three methods (hereinafter referred to as “embedded information”), first, the above-described states (1) to (4) are obtained by extracting the embedded method identification information. Is determined. Subsequently, the embedded information is extracted by an extraction process corresponding to the determined state. Therefore, in the above cases (1), (2), and (4), execution of unnecessary processing is suppressed, and efficient embedded information extraction processing becomes possible.

  Note that the method used to embed the embedded information can embed an information amount of several tens to several hundreds of bits. Such a method generally requires “non-real-time processing” when extracting information. “Non-real-time processing” means, for example, when copying a document on which an image with embedded information is printed, output processing is performed until the entire image or the amount of image information necessary for information extraction is captured in the frame memory. The process that can be stopped. An information embedding method that requires “non-real-time processing” when extracting information is referred to as a “non-real-time method” in the present embodiment. In general, a method capable of embedding information amount of several tens to several hundreds of bits corresponds to a non-real time method.

  All of the three types of methods in the present embodiment correspond to non-real time methods. However, the three types of methods in the present embodiment are merely examples, and the embedded information may be embedded by other methods. Further, the embedded information may be embedded by two types or four or more types. Further, when embedded information is embedded simultaneously by each method (that is, the state of (2) or (3) above), the embedded information may be the same information or different information.

  On the other hand, the method for embedding the embedding method identification information only needs to be able to embed the information amount of the number of bits (about several bits) that can identify the type of embedding method. There is a certain degree of correlation between the amount of information that can be embedded and the performance of information extraction processing for the information amount. That is, the smaller the amount of information that can be embedded, the simpler the information extraction process, and the “non-real-time process” becomes unnecessary. In other words, information extraction processing can be performed in parallel while scanning an image line by line with a line sensor. Therefore, in the present embodiment, the embedding method identification information is embedded by a method that embeds less information than any of the above three methods, that is, by a method that does not require non-real-time processing. A method that does not require “non-real time processing” is referred to as “real time method” in the present embodiment.

  That is, in the present embodiment, information for embedding embedded information in a plurality of non-real-time methods and identifying information used for embedding embedded information can be embedded more than any of the plurality of non-real-time methods. The embedding information extraction process is made more efficient by embedding it in a small amount in real time.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. A method for embedding embedding method identification information will be described. In the present embodiment, the embedding method identification information is embedded in the image by synthesizing (superimposing) a pattern composed of a plurality of dots (hereinafter referred to as “dot pattern”) on the background of the image.

  For example, the following dot patterns are used. FIG. 1 is a diagram showing an example of a dot pattern used in the embodiment of the present invention. In FIG. 1, a dot pattern a is composed of three dots each having a relative positional relationship. In implementing the present invention, the number of dots constituting one dot pattern may be three or more. Further, the pattern is not necessarily formed by dots. For example, a pattern constituted by a line segment or a combination of a line segment and a dot may be used.

  FIG. 2 is a diagram illustrating an example of an image in which dot patterns are combined. FIG. 2 shows an example in which the dot pattern a is repeatedly combined with the image 500. In the figure, the dot pattern a is shown enlarged so that it can be easily seen, but actually, it is configured as a finer pattern.

  By synthesizing the dot pattern a with the image, at least one bit of information can be embedded in the image 500. That is, there are a state where the dot pattern a is synthesized and a state where it is not synthesized. However, one bit is not enough to identify the states (1) to (4). Strictly speaking, the state (1) is further classified into three states if the method adopted as any one method is distinguished. Similarly, the state of (2) is further classified into three states, if the methods adopted as any two methods are distinguished. Therefore, it is necessary to distinguish a total of eight states, and for this purpose, an information amount of at least 3 bits is required as embedding method identification information.

  Therefore, in the present embodiment, a meaning is given to the relative angle difference between the dot patterns a. That is, a dot pattern a or a pattern obtained by rotating the dot pattern a by an arbitrary angle (hereinafter referred to as “basic pattern”) and a dot pattern obtained by rotating the basic pattern by a predetermined angle (hereinafter referred to as “additional pattern”). The embedding method identification information is expressed by a relative angle. Here, the center of rotation when rotating the dot pattern a is not limited to a specific position, but it is necessary to be consistent with the center of rotation when extracting information. In the present embodiment, information is extracted using a pattern dictionary described later in FIG. In the pattern dictionary of FIG. 17, a pixel located at the center coordinate in a rectangular area including the dot pattern a (when the width of the rectangular area is W and the height is H, it is located at (W / 2, H / 2). Pixel) is the center of rotation. Therefore, in the present embodiment, the center of rotation when embedding information is also a pixel located at the center coordinate in the rectangular region including the dot pattern a.

  FIG. 3 is a diagram illustrating examples of basic patterns and additional patterns. FIG. 3 shows an example in which the basic pattern a and the additional pattern b have an angle of 45 degrees relatively. That is, the additional pattern b in the figure is obtained by rotating the basic pattern a 45 degrees clockwise.

  FIG. 4 is a diagram illustrating an example of an image in which information is embedded according to the relative angle of the basic pattern and the additional pattern. In FIG. 4, the drawing elements (the house picture) of the image 500 are omitted for convenience. Also, the arrows in the drawing are for indicating the direction (angle) of each dot pattern, and are not drawing elements on the image. FIG. 4 shows an example in which two basic patterns a and a plurality of additional patterns b obtained by rotating the basic pattern a 45 degrees clockwise are combined. In FIG. 4, only two are combined for convenience, but considering practicality, it is desirable that a larger number of dot patterns be combined as will be described later. The number of basic patterns a and the number of additional patterns b are preferably the same or substantially the same from the viewpoint of detection accuracy. In this method, the absolute or relative arrangement position of the basic pattern a and the additional pattern b is not questioned.

  Since the relative angle is a problem, it does not matter which is the basic pattern and which is the additional pattern in the embedding method identification information embedding process and extraction process. That is, in this method, the names of the basic pattern and the additional pattern are merely convenient.

  According to the method of embedding information according to the relative angle of the two dot patterns, it is possible to embed information of about several bits by changing the relative angle of both patterns. Since the maximum value of the relative angle between the basic pattern a and the additional pattern b is 180 degrees, for example, if quantization is performed in 8 steps by 22.5 degrees, 3-bit information can be embedded. In this embodiment, among the three types of methods for embedding embedded information, consideration is given to the feasibility of a method for embedding embedded information using a dot pattern (details will be described later). The state where the relative angle is 0 degrees is not used for embedding the embedding method identification information. Accordingly, when quantization is performed in increments of 22.5 degrees, the value that can be taken as the relative angle is 22.5 × n (1 ≦ n ≦ 8).

  In the present embodiment, embedding method identification information is assigned to each relative angle quantized by 22.5 degrees as shown in FIG.

  FIG. 5 is a diagram illustrating the correspondence between the embedding method identification information and the relative angle. In the table shown in FIG. 5, the values of 000, 001, 010, 011, 100, 101, 110, and 111 are set according to the value of n for a relative angle of 22.5 × n (1 ≦ n ≦ 8). Assigned. By this value, the embedding method of embedding information in the image (which method is used to embed the embedding information) is identified. The reason why the method of embedding information with respect to the relative angle between the two dot patterns corresponds to the real-time method will be described later.

  Next, three types of non-real time methods for embedding embedded information will be described. In this embodiment, an example in which a barcode method, an information embedding method based on character outline shape deformation, and an information embedding method based on arrangement information of dot patterns with different angles are planned as candidates for embedding information is exemplified. And

  The barcode method is a method of embedding information by combining a barcode image representing embedded information with a part of an image (for example, one of four corners). The barcode may be a two-dimensional barcode. Hereinafter, the barcode method is also referred to as “non-real-time first method”.

  The information embedding method by deforming the shape of the character outline is an effective method for an image including characters as drawing elements, and is a method of embedding information by deforming the outline of the character. For this method, “Digital Watermarking Method for Lettering Using Character Shape” (November 25, 1999, Shuichi Sugoi, Masayoshi Kamijo, published by The Institute of Electronics, Information and Communication Engineers, pages 2175-2177) Be familiar with. Hereinafter, the information embedding method by deforming the character outline portion is also referred to as “non-real-time second method”.

  An information embedding method based on arrangement information of dot patterns having different angles will be described. In this method, the relative arrangement information (in other words, the basic pattern a and the additional pattern b) in the method used for embedding the embedding method identification information (that is, the method of embedding information with respect to the relative angle of two dot patterns). This is a method of embedding information in an array composed of a basic pattern a and an additional pattern b. Hereinafter, this method is also referred to as “non-real-time third method”.

  FIG. 6 is a diagram illustrating an example of an image in which information is embedded by an arrangement of basic patterns and additional patterns. FIG. 6 shows an example in which the array length is 4, and when the array is scanned in order from left to right and from top to bottom, the basic pattern a, the additional pattern b, the basic pattern a, and the additional pattern b are sequentially displayed. A dot pattern is arranged. For example, if “0” is assigned to the basic pattern a and “1” is assigned to the additional pattern b, the value “0110” is embedded in the image 500 of FIG. Therefore, in this method, the basic pattern a and the additional pattern b are clearly distinguished. However, in this method, the relative angle between the basic pattern a and the additional pattern b is not questioned. It is sufficient that there is an identifiable angle difference between the two dot patterns, and the embedded information is not affected no matter what the angle difference is.

  Also in this method, in consideration of information extraction accuracy, it is desirable that the same number of basic patterns a and additional patterns b are embedded in one array. Then, when the array length is “4” as shown in FIG. 6, the embedding information that can be embedded by this method is as shown in FIG.

  FIG. 7 is a diagram illustrating a correspondence between an array and embedded information when the array length is 4. As shown in FIG. 7, when the arrangement length is 4, under the restriction that the same number of basic patterns a and additional patterns b are provided, six types of embedding information can be embedded. Therefore, an information amount of more than 2 bits can be embedded. However, FIG. 6 shows an example in which the arrangement length has become extremely small due to the enlarged dot pattern for easy understanding. In practice, since a lot of finer dot patterns can be embedded on the entire surface of the image, the arrangement length also becomes longer. Therefore, information of about 100 bits can be embedded. In FIG. 7, the value of the embedded information is simply a numerical value (0 to 5), but what information is assigned to each array is arbitrary. If it is about 100 bits, the document ID of the image can be assigned.

  By the way, in the present embodiment, a method for embedding embedding method identification information is to synthesize a dot pattern with the entire background of an image. Therefore, interference with a method for embedding embedded information (that is, a situation in which a barcode or a character overlaps with a dot pattern) may occur, which may hinder extraction of embedded information. In order to avoid such interference, it is preferable not to synthesize a dot pattern for the barcode area or the character peripheral area.

  FIG. 8 is a diagram illustrating a first example in which interference between the barcode area or the character peripheral portion and the dot pattern is avoided. In the example of FIG. 8, no dot pattern is synthesized around the barcode b1 or the character c1. By doing so, interference between the barcode or character outline and the dot pattern is avoided, and an extraction error when extracting information from the barcode or character outline can be prevented. Such a technique is described in detail in JP-A-2006-229924.

  Further, a dot pattern may be synthesized as shown in FIG. FIG. 9 is a diagram illustrating a second example in which interference between the barcode area or the character peripheral portion and the dot pattern is avoided. FIG. 9 shows an example in which a dot pattern is synthesized only in the peripheral portion of the image (the blank portion when printed). Even with such a configuration, it is possible to effectively avoid interference due to the dot pattern.

  As for the method for embedding information by the dot pattern arrangement, the same dot pattern as that for embedding by the relative angle of the dot pattern may be used. By doing so, information can be embedded twice by the same dot pattern, and interference by both methods can be avoided. For example, referring to FIGS. 5 and 7, it is necessary to embed embedding method identification information “001” according to the relative angle between the basic pattern a and the additional pattern b, and “1” is embedded according to the arrangement of both patterns. If it is necessary to embed the pattern, the relative angle between the basic pattern a and the additional pattern b may be 45 degrees, and the basic pattern a and the additional pattern b may be arranged in the order shown in FIG. By doing so, when extracting information, the relative angle between the basic pattern a and the additional pattern b is detected as 45 degrees, and the arrangement of both patterns is detected as shown in FIG. Therefore, information is appropriately extracted from any method. In the method of embedding information in the dot pattern arrangement, the relative angle of both dot patterns is not questioned. In the method of embedding information with respect to the relative angle of the dot pattern, the arrangement position of both dot patterns is not questioned. This is because of compatibility in both directions.

  In the present embodiment in which the three types of non-real-time methods described above are scheduled as candidates for the method for embedding embedded information, the embedding method identification information and the embedded state of the embedded information in the image are as follows: Associate.

  FIG. 10 is a diagram showing a correspondence table between the embedding method identification information and the embedding state of embedding information. In FIG. 10, a non-real-time first method, a non-real-time second method, and a non-real-time third method are allocated in order from the upper bit of the 3-bit embedding method identification information. When the bit value is “1”, it indicates that the embedded information is embedded by a method corresponding to the bit. When the bit value is “0”, it indicates that the embedded information is not embedded depending on the method corresponding to the bit. Therefore, for example, when the value of the embedding method identification information is “000”, it indicates that the embedding information is not embedded by any method. When the value of the embedding method identification information is “100”, it indicates that the embedding information is embedded only by the non-real-time first method. When the value of the embedding method identification information is “111”, it indicates that the embedding information is embedded by all three types of non-real time methods.

  Hereinafter, an apparatus and the like for realizing the method described above will be described.

  FIG. 11 is a diagram illustrating a configuration example of an image processing apparatus that executes an information embedding process. The image processing apparatus 10 can be constructed by, for example, a general-purpose computer, a printer or a multi-function machine capable of mounting an application. In FIG. 11, the image processing apparatus 10 includes an image data acquisition unit 101, an information input unit 102, a first dot pattern generation unit 103, a second dot pattern generation unit 104, a barcode generation unit 105, a character outline deformation unit 106, information It comprises an embedding control unit 107, a selector 108, a composition processing unit 109, a printing unit 110, and the like.

  The image data acquisition unit 101 acquires or generates data (target image data) including an image (hereinafter referred to as “target image”) that is an information embedding target. For example, the image data acquisition unit 101 may be configured by word processing software that generates document data, a program that converts document data generated by the word processing software into image data, or the like, or an image stored in advance. Data reading processing may be executed. The information input unit 102 selects information to be embedded in the target image (embedding information) and information indicating whether to embed the embedded information by selecting one of the non-real-time first to third methods (hereinafter referred to as “embedding method selection information”). ")."

  The first dot pattern generation unit 103 receives embedding method selection information as input information. The first dot pattern generation unit 103 determines the value of the embedding method identification information according to the embedding method selection information, and images of two dot patterns (basic pattern a and additional pattern a) having a relative angle representing the value. Data (dot pattern data) is generated and output. For determining the value of the embedding method identification information according to the embedding method selection information, for example, a correspondence table between the embedding method selection information and the embedding method identification information is registered in the storage device of the image processing apparatus 10, and the correspondence table is displayed. Based on this.

  The second dot pattern generation unit 104 receives embedding method selection information and embedding information as input information. The second dot pattern generation unit 104 determines the value of the embedding method identification information according to the embedding method selection information, and determines the relative angle that expresses the value. The second dot pattern generation unit 104 also determines the arrangement of dot patterns that represent the embedded information. The second dot pattern generation unit 104 further generates and outputs dot pattern data so that the basic pattern a having the determined relative angle and the additional pattern b form the determined array.

  The barcode generation unit 105 accepts embedded information as input information. The barcode generation unit 105 generates and outputs barcode image data (barcode data) according to the embedded information.

  The character outline deforming unit 106 receives target image data and embedded information as input information. The character outline deforming unit 106 deforms the outline of the character included in the target image according to the embedded information. The character outline deforming unit 106 outputs target image data in which the character outline is deformed.

  The information embedding control unit 107 selects the embedding method of embedding information by controlling the selector 108 according to the embedding method selection information received by the information input unit 102. That is, the information embedding control unit 107 and the selector 108 perform selection of data output from the first dot pattern generation unit 103, the second dot pattern generation unit 104, the barcode generation unit 105, or the character outline transformation unit 106. Is called.

  The composition processing unit 109 generates image data in which the embedding information is embedded by combining the data selected by the selector 108 with the target image. The printing unit 110 controls printing of image data generated by the synthesis processing unit 109 on a paper medium (printing paper).

  In addition, each said part may be comprised by hardware (circuit), and may be comprised by software. In the case of being configured by hardware, for example, the first dot pattern generation unit 103, the second dot pattern generation unit 104, the barcode generation unit 105, and the character outline transformation unit according to the input of the embedding information and the embedding method selection information The processing by 106 is executed in parallel. The data output from these units is selected by the selector 108 under the control of the information embedding control unit 107 based on the embedding method selection information, and is output to the synthesis processing unit 109.

  For example, when the non-real-time first method is specified in the embedding method selection information, the barcode data from the barcode generation unit 105 and the dot pattern data from the first dot pattern generation unit 103 are combined. 109 is output. The synthesis processing unit 109 synthesizes the barcode data and the dot pattern data with the target image data.

  When the non-real-time second method is designated, the target image data from the character outline deformation unit 106 and the dot pattern data from the first dot pattern generation unit 103 are output to the synthesis processing unit 109. The composition processing unit 109 synthesizes the dot pattern data with the target image data from the character outline deforming unit 106. That is, when the non-real-time second method is designated, the target image output from the character outline deforming unit 106 is a target to be synthesized such as a barcode or a dot pattern.

  When the non-real-time third method is designated, the dot pattern data from the second dot pattern generation unit 104 is output to the synthesis processing unit 109. The composition processing unit 109 synthesizes the dot pattern with the target image data.

  If it is specified that no embedding information is embedded by any of the methods, the dot pattern data from the first dot pattern generation unit 103 is output to the synthesis processing unit 109.

  On the other hand, when each part of FIG. 11 is implement | achieved by software, the program concerning the said software makes CPU, such as the said computer, perform the process as shown in FIG. The program may be recorded in advance in the ROM of the device, or may be downloaded via a network. Further, it may be installed from a recording medium 600 such as a CD-ROM.

  FIG. 12 is a flowchart for explaining information embedding processing by the image processing apparatus.

  In step S <b> 101, the image data acquisition unit 101 acquires an image (target image) that is an information embedding target, and develops the target image on the memory of the image processing apparatus 10. Subsequently, the information input unit 102 receives input of embedding information and embedding method selection information (S102). For example, the embedded information may be input on the input screen by displaying the input screen on the display device. When the embedding information is not input (No in S103), the first dot pattern generation unit 103 determines that the value of the embedding method identification information is “000”, and the relative angle corresponding to “000” (22. The dot pattern data (first dot pattern) of the basic pattern a and the additional pattern b having 5 degrees is generated (S104). Subsequently, the synthesis processing unit 18 synthesizes the dot pattern data with the target image (S105). The printing unit 110 causes the printer to print the target image (S114).

  On the other hand, when the embedded information is input (Yes in S103), the embedded information control unit 107 determines whether the non-real-time first method is selected based on the embedded method selection information (S106). When the non-real-time first method is selected (Yes in S106), the barcode generation unit 105 generates barcode data corresponding to the embedded information (S107). The composition processing unit 109 synthesizes the barcode data with the target image (S108). If the non-real-time first method is not selected (No in S106), steps S107 and S108 are not executed.

  Subsequently, the embedded information control unit 107 determines whether the non-real-time second method is selected based on the embedded method selection information (S109). When the non-real-time second method is selected (Yes in S109), the character outline deforming unit 106 deforms the outline of the character in the target image according to the embedded information (S110). If the non-real-time second method is not selected (No in S109), step S110 is not executed.

  Subsequently, the embedding information control unit 107 determines whether or not the non-real-time third method is selected based on the embedding method selection information (S111). When the non-real-time third method is selected (Yes in S111), the second dot pattern generation unit 104 generates dot pattern data (second dot pattern) according to the embedding information and the embedding method selection information ( S112). The composition processing unit 109 synthesizes the dot pattern data with the target image (S113). Subsequently, the printing unit 110 causes the printer to print the target image (S114). If the non-real-time third method is not selected (No in S111), steps S104 and S105 are executed instead of steps S112 and S113. The dot pattern generated by the second dot pattern generation unit 104 also includes embedding method identification information. If the dot pattern generated by the second dot pattern generation unit 104 is not synthesized, the first dot pattern This is because the embedding method identification information is not embedded in the target image unless the dot pattern generated by the generation unit 103 is synthesized.

  In the information embedding process, if information is not embedded by any non-real time method, the dot pattern indicating the embedding method identification information may not be combined with the target image.

  Next, information extraction will be described. FIG. 13 is a diagram illustrating a hardware configuration example of an image processing apparatus that executes information extraction processing. 13, the image processing apparatus 20 includes a scanner 21, a RAM 22, a DSP (Digital Signal Processor) 23, a plotter 24, an embedding method determination unit 25, a controller 26, an HDD (Hard Disk Drive) 27, an operation panel 28, and the like. .

  The scanner 21 reads an image from the document 700. The read image data is input to the RAM 22. The RAM 22 implements a memory area that is used as a line memory that outputs input image data by the FIFO method. The line memory has a capacity capable of storing, for example, about several tens of lines from the viewpoint of speeding up image processing and reducing the cost of the image processing apparatus 10. That is, only part of the image data can be stored (expanded) in the line memory. The DSP 23 is a processor that executes various kinds of image processing (for example, background removal, gamma correction, gradation conversion, etc.) on the image data. The embedding method determination unit 25 is a circuit that detects a dot pattern from image data and extracts embedding method identification information based on a relative angle difference between the dot patterns. However, the embedding method determination unit 25 may be configured by software. In this case, the program of the software is loaded into the RAM 22 and processed by the DSP 23, whereby the function of the embedding method determination unit 25 is realized.

  The controller 26 includes a CPU 261, a RAM 262, a ROM 263, and the like, and image processing (information extraction processing) that cannot be processed by the RAM 22 and the DSP 23 or the like due to memory consumption, processing complexity, complexity, or the like. And a process for controlling other functions of the image processing apparatus 20 is executed. The ROM 263 stores a program for executing the processing. The program is loaded into the RAM 262 at the time of execution. The RAM 262 is also used as a frame memory for storing image data for image processing executed in the controller 26. The capacity of the frame memory needs to be at a level that can store at least the amount of image data necessary for the image processing, for example, a level that can store the entire image read from the document 700. Therefore, the capacity of the frame memory is at least larger than the capacity of the line memory in the RAM 22.

  The CPU 261 executes the processing by executing the program loaded in the RAM 262. For example, document management information is registered in the HDD 27. Here, the document management information refers to information configured by document ID, attribute information, and the like for each document. The operation panel 28 is composed of a liquid crystal panel, buttons, and the like, and accepts user operations.

  The document 700 is a document printed by the image processing apparatus 10.

  By the way, the components of the image processing apparatus 20 are distinguished from the part that executes “non-real-time processing” in this embodiment and the part that does not, and the former is called “non-real-time processing unit”, and the latter For convenience, the scanner 21, RAM 22, DSP 23, plotter 24, and embedded information determination unit 25 are included in the real-time processing unit. On the other hand, the controller 26 and the HDD 27 are included in the non-real-time processing unit.

  Software implemented by a program for causing the CPU 261 to execute an embedded information extraction function (information extraction function) will be described. FIG. 14 is a diagram illustrating a software configuration example of the information extraction function. In FIG. 14, the information extraction function executed by the CPU 261 includes a barcode information extraction unit 261a, a character outline information extraction unit 261b, a dot pattern information extraction unit 261c, and an extraction information processing unit 261d.

  The barcode information extraction unit 261a extracts information embedded by the non-real-time first method from the image. That is, the barcode information extraction unit 261a detects a barcode combined with an image and extracts embedded information from the barcode. The character outline information extraction unit 261b extracts information embedded by the non-real-time second method from the image. That is, the character outline information extraction unit 261b extracts embedded information from the outlines of characters included in the image. The dot pattern information extraction unit 261c extracts information embedded by the non-real-time third method from the image. That is, the dot pattern information extraction unit 261c detects a dot pattern combined with an image, and extracts embedded information from the dot pattern arrangement. The extracted information processing unit 261d executes processing control according to the value of the extracted embedded information.

  Hereinafter, the processing procedure of the image processing apparatus 20 in FIG. 13 will be described. FIG. 15 is a flowchart for explaining information extraction processing by the image processing apparatus. FIG. 15 illustrates an example in which the embedding information for the document 700 is a document ID, and copy control of the document 700 is performed based on the document ID. Here, the document ID is a document ID on a computer system (for example, a document management system) of image data printed on the original 700. In this computer system, it is assumed that access control information (e.g. whether copying is possible) is managed for each document.

  When an instruction to copy the original 700 is input from the operation panel 28, the scanner 21 of the image processing apparatus 20 reads an image of the original 700 (hereinafter referred to as “target image”) (S201). The target image is written (stored) in the line memory of the RAM 22 and the frame memory of the RAM 262 as needed for the read lines (S202, S203).

  When the target images for the number of lines in the line memory are written into the line memory, the embedding method determination unit 25 detects the dot pattern (basic pattern a and additional pattern b) from the target image read into the line memory, and the basic pattern The embedding method identification information is extracted by determining the relative angle between a and the additional pattern b (S204). Further, in the real-time processing unit, image processing by the DSP 23 is executed on the target image in the line memory before and after the processing by the real-time method determining unit 25.

  By the way, the contents of the image in the line memory change by one line or several lines so that when a new line is input, the oldest line is output. The processing of the real-time processing unit is executed in real time as needed according to changes in the contents of the line memory. From the viewpoint of detection accuracy, it is desirable that the detection of the dot pattern by the embedding method determination unit 25 is performed every time the contents of the line memory change. However, as will be described later, the embedding method identification information can be extracted before the reading of all lines of the image of the original 700 is completed.

  When the embedding method identification information is extracted (Yes in S205), the embedding method determination unit 25 determines whether the embedding information is embedded in the target image based on the value of the embedding method identification information (S206). This determination is made based on the table shown in FIG. Note that information corresponding to the table shown in FIG. 10 is recorded in the storage device of the image processing apparatus 20.

  If the value of the embedding method identification information is other than “000” (Yes in S206), the embedding method determination unit 25 instructs the plotter 24 to wait for output processing (S207). The real-time processing unit can start output processing by the plotter 24 in response to completion of image processing by the DSP 23 even if not all lines of the image of the original 700 have been read. This is because copying such a document may be prohibited. Subsequently, the embedding method discriminating unit 25 extracts the embedding information by a method according to the value of the embedding method identification information, and performs the information extraction function (the barcode information extracting unit 261a, the character outline information extracting unit 261b, and the dot pattern) of the controller 26. Instruct to at least one of the information extraction units 261c).

  Receiving the embedded information extraction instruction, the controller 26 waits for processing until the target images for the number of lines sufficient to extract the embedded information are written in the frame memory of the RAM 262. With regard to any of the barcode, character outline, and dot pattern arrangement, it is affected by the direction in which the target image is read. Basically, it waits until almost all lines of the target image are written in the frame memory.

  When the target images for a sufficient number of lines are stored in the frame memory, the barcode information extraction unit 261a, the character outline information extraction unit 261b, and the extraction unit instructed to extract the embedded information among the dot pattern information extraction unit 261c Extracts embedded information (document ID) based on the respective methods (S209). Subsequently, the extracted information processing unit 261d determines whether or not the original 700 can be copied based on the extracted document ID (S210). For example, the extracted information processing unit 261d acquires security information for the document ID from a document management system built in the HDD 27 or a computer connected via the network, and determines whether or not copying is possible based on the security information. (S211).

  When copying is permitted (Yes in S211), the extraction information processing unit 261d cancels standby for output in the plotter 24. As a result, a copy of the original 700 is printed (output) on the printing paper by the plotter 24 (S212). On the other hand, when copying is prohibited (No in S211), the extraction information processing unit 261d instructs the plotter 24 to stop output (S213). As a result, the copying of the original 700 is stopped. The extracted information processing unit 261d may instruct the DSP 23 to execute image processing such as painting an output image. Then, if the output standby of the plotter 24 is canceled, an image in which the image of the original 700 is filled is output to the printing paper, and the execution of copying is substantially avoided.

  By the way, when the embedding method identification information is not extracted from the image of the document 700 (No in S205) or when the value of the embedding method identification information is “000” (No in S206), the embedding method determination unit 25 No output standby instruction for the plotter 24 or embedded information extraction instruction for the controller 26 is performed. Therefore, in this case, the plotter 24 outputs an image without waiting for extraction of embedded information. Therefore, a copy of the original 700 is printed on the printing paper from the plotter 24 with the same performance as in normal copying (S212). Here, normal copying means copying when the embedded information extraction function is disabled. The reason why copying can be executed with the same performance as in normal copying is that the influence of the extraction process of the embedding method identification information by the embedding method determination unit 25 on the copying process is small.

  The process of extracting embedding method identification information will be described. FIG. 16 is a flowchart for explaining the process of extracting embedding method identification information by the embedding method discrimination unit.

  First, a dot pattern is detected from the line memory (S2041). The dot pattern is detected by, for example, pattern matching using a pattern dictionary registered in the storage device of the image processing apparatus 10.

  FIG. 17 is a diagram illustrating an example of a pattern dictionary. FIG. 17 shows an example of a pattern dictionary composed of 16 patterns obtained by rotating the dot pattern a by 22.5 degrees. Hereinafter, each pattern constituting the pattern dictionary is referred to as a “master pattern”. Note that FIG. 17 shows an example in which the center coordinates of the rectangular area including the dot pattern a is the center of rotation.

  On the other hand, for example, several lines of the target image are stored in the line memory as follows. FIG. 18 is an image diagram of a line memory. In FIG. 18, two line patterns are included in the line memory 22L. Pattern matching is performed on such an image in the line memory 22L. Therefore, the number of lines in the line memory 22L needs to be at least the number of dots in the vertical and horizontal directions of the dot pattern to be detected. FIG. 17 shows an example of 15 × 15 dots. In this case, the line memory 22L needs a capacity of 15 lines or more.

  Pattern matching is performed for each dot pattern included in the pattern dictionary while shifting the line memory 22L pixel by pixel. Each time a pattern is matched (a dot pattern is detected), the total number of detections and the number of detections for each angle of the dot pattern are incremented (S2042). The angle of the dot pattern is determined by the angle of the master pattern at the time of pattern detection. Such pattern matching is executed every time the contents of the line memory 22L are updated until the total number of detections exceeds a preset threshold (No in S2042). Note that the threshold is provided in order to prevent erroneous determination due to a pattern that happens to match.

  When the total number of detections exceeds the threshold, the detection of the dot pattern a ends (Yes in S2042), and two peaks are detected for the number of detections for each angle (S2043). In other words, the top two angles with the largest number of detections are determined (S2045). When two peaks are detected (Yes in S2046), the value of the embedding method identification information is determined based on the angle difference between the two peaks (S2047). That is, the angle difference corresponds to a relative angle between the basic pattern a and the additional pattern b. The value of the embedding method identification information is determined based on the table shown in FIG. That is, a value corresponding to the angle difference is retrieved from the table shown in FIG. Note that information corresponding to the table shown in FIG. 5 is registered in advance in the storage device of the image processing apparatus 10, for example.

  The dot pattern detection is stopped if the threshold value is not exceeded even after a predetermined number of lines have been detected (Yes in S2044). In this case, it is determined that the embedding method identification information is not embedded (S2048). The predetermined line may be set to a value that does not hinder real-time processing.

  The dot pattern a is combined with the background of the image of the original 700 (the entire surface of the image, or at least the surrounding margin). Therefore, according to the processing shown in FIG. 16, the embedding method identification information can be extracted without being influenced by the direction in which the document 700 is set. Therefore, the extraction of the embedding method identification information is sufficient with the line memory. Further, the extraction process of the embedding method identification information is a process executed based on the image in the line memory, so that the influence on the performance of the entire copying process can be suppressed to a small level. Therefore, the embedding method identification information can be extracted in the process of performing the copying process (that is, in real time).

  Note that the administrator can set the information extraction function in the image processing apparatus 20 to be invalid when it is desired to copy any document regardless of whether or not the embedded information in the non-real-time method is embedded. It is good.

  By the way, in this embodiment, the example in which the document ID is embedded as the embedded information has been described. This is because it is generally difficult to embed document management information such as a printer, a printing device, and a printing date and time in an information amount of about 100 bits. However, such circumstances are considered to differ depending on the environment. For example, in a small office or the like, the amount of managed documents is small, and it is possible to directly embed information such as a printer, a printing device, and a printing date and time. In that case, the cost for preparing a document management system or the like can be reduced. The “print date / time” can be used, for example, for the control of prohibiting copying within three months from the date / time of printing.

  In the present embodiment, a case has been described in which a plurality of non-real time systems are different systems. However, when only one embedding method is used and a plurality of types of configurations of information embedded by the embedding method are allowed, the embedding method identification information in the present embodiment is used as information for identifying the configuration of the information. May be used.

  For example, when error correction coding is performed on embedded information, the error correction capability strength parameter (or the amount of information that can be embedded) can be selected from a plurality of candidates. Is given as an example. In this case, each time the error correction strength parameter is different, the information structure is different, and each can be regarded as a different embedding method.

  Actually, when information is embedded in a document image with a dot pattern, information cannot be embedded in a portion overlapping with content such as characters, figures, and photographs. Further, even when information is embedded in a character outline shape, the embedded information cannot always be extracted with 100% accuracy due to the influence of noise mixing during printing. Even in view of such circumstances, error correction coding of embedded information is effective. The appropriate level of error correction strength varies depending on the contents of the document image and the like, and it is also meaningful to be able to select the strength. In this case, only one parameter (embedding method in the present embodiment) can be selected. At the time of information extraction, an error correction code information extraction process is performed according to the value of this parameter.

  Specifically, when the (7, K) Reed-Solomon code is used for error correction, it is possible to use such that one of K values can be selected from 1, 3, and 5. K is a parameter indicating how many symbols among all symbols 7 including those for error correction are allocated to symbols for original information. When K = 3, 1- (3/7) of the total embeddable information amount when error correction is not performed is allocated for error correction. Further, 3/7 of the total embeddable information amount when error correction is not performed is assigned to the original information symbol. Therefore, the amount of information that can be embedded is 3/7 of the case where error correction is not performed.

  As described above, according to the present embodiment, information is embedded in an image (including a document image) on a medium such as paper by selecting one or a plurality of methods from a plurality of non-real time methods. When performing, ID information (embedding method identification information) indicating which method is embedded is embedded in the image by a method capable of extracting information in real time. Therefore, when the embedded information is extracted, it is possible to avoid unnecessary non-real time information extraction processing, and it is possible to prevent the performance of the information extraction processing from being deteriorated.

  Therefore, when the document ID is embedded as the embedded information and copy control (copy permission / denial control) is performed based on the document ID, it is possible to prevent the performance of the document ID extraction process from being deteriorated. As a reflective effect, even for documents that do not require copy prohibition control, it is determined that the document ID is not embedded by real-time information extraction processing, so unnecessary execution of information extraction processing by a non-real-time method is avoided. can do. Therefore, the occurrence of a situation in which the normal copying process is interrupted and the user is kept waiting is avoided, and a decrease in productivity of the normal copying process is prevented.

  As mentioned above, although the Example of this invention was explained in full detail, this invention is not limited to such specific embodiment, In the range of the summary of this invention described in the claim, various deformation | transformation・ Change is possible.

It is a figure which shows the example of the dot pattern used in embodiment of this invention. It is a figure which shows the example of the image with which the dot pattern was synthesize | combined. It is a figure which shows the example of a basic pattern and an additional pattern. It is a figure which shows the example of the image in which information was embedded by the relative angle of the basic pattern and the additional pattern. It is a figure which shows a response | compatibility with embedding system identification information and a relative angle. It is a figure which shows the example of the image in which information was embedded by the arrangement | sequence of a basic pattern and an additional pattern. It is a figure which shows a response | compatibility with the arrangement | sequence when a sequence length is 4, and embedding information. It is a figure which shows the 1st example which avoided the interference with a barcode area | region or a character peripheral part, and a dot pattern. It is a figure which shows the 2nd example which avoided interference with a barcode area | region or a character peripheral part, and a dot pattern. It is a figure which shows the correspondence table of embedding method identification information and the embedding state of embedding information. It is a figure which shows the structural example of the image processing apparatus which performs the embedding process of information. It is a flowchart for demonstrating the embedding process of the information by an image processing apparatus. It is a figure which shows the hardware structural example of the image processing apparatus which performs the extraction process of information. It is a figure which shows the software structural example of an information extraction function. It is a flowchart for demonstrating the information extraction process by an image processing apparatus. It is a flowchart for demonstrating the extraction process of the embedding system identification information by an embedding system discrimination | determination part. It is a figure which shows the example of a pattern dictionary. It is an image figure of a line memory.

Explanation of symbols

10, 20 Image processing device 21 Scanner 22 RAM
23 DSP
24 Plotter 25 Embedding Method Discriminator 26 Controller 26
27 HDD
28 Operation Panel 101 Image Data Acquisition Unit 101
102 information input unit 103 first dot pattern generation unit 104 second dot pattern generation unit 105 bar code generation unit 106 character outline deformation unit 107 information embedding control unit 108 selector 109 synthesis processing unit 110 printing unit 261 CPU
261a Barcode information extraction unit 261b Character outline information extraction unit 261c Dot pattern information extraction unit 261d Extraction information processing unit 262 RAM
263 ROM
600 Recording medium 700 Original

Claims (22)

An image processing method in which an image processing apparatus embeds information in an image,
An information embedding procedure for embedding the information using one or more methods selected from among a plurality of methods for embedding information in the image;
A method identification information embedding procedure for embedding method identification information for identifying the one or more selected methods in the image;
The method identification information embedding procedure includes embedding the method identification information by a method that embeds less information than any method used in the information embedding procedure.   The image processing method according to claim 1, wherein the method identification information embedding procedure combines a predetermined pattern with a background of the image.   3. The image according to claim 2, wherein the scheme identification information embedding procedure combines the predetermined pattern and a pattern obtained by rotating the predetermined pattern in accordance with a value of the scheme identification information with the image. Processing method.   4. The image processing method according to claim 2, wherein the predetermined pattern is a pattern composed of a plurality of dots. An image processing method in which an image processing apparatus extracts information from an image in which information may be embedded by a plurality of methods,
A method identification information extraction procedure for extracting method identification information for identifying a method used for embedding information from the image;
An information extraction procedure for extracting information embedded in the image using one or more methods identified by the method identification information;
The image processing method, wherein the method identification information is embedded in the image by a method that embeds less information than any method used in the information extraction procedure. The image processing apparatus includes a first storage device capable of storing a part of the image, and a second storage device capable of storing the entire image.
An image reading procedure for reading the image from a medium;
A storage procedure for storing the read image in the first storage device and the second storage device, wherein the method identification information extraction procedure is performed from the image stored in the first storage device; Extract method identification information,
The image processing method according to claim 5, wherein the information extraction procedure extracts the information from an image stored in the second storage device.   7. The image processing method according to claim 5, wherein the information extraction procedure is not executed according to the value of the method identification information.   The image processing method according to claim 7, wherein the method identification information extraction procedure extracts the method identification information based on a predetermined pattern combined with a background of the image.   9. The method identification information according to claim 8, wherein the method identification information extraction procedure determines a value of the method identification information based on an angle difference between the predetermined pattern and a pattern obtained by rotating the predetermined pattern. Image processing method.   The image processing method according to claim 8, wherein the predetermined pattern is a pattern composed of a plurality of dots. An image processing apparatus for embedding information in an image,
Information embedding means for embedding the information using one or more methods selected from a plurality of methods for embedding information in the image;
A method identification information embedding unit for embedding method identification information for identifying the one or more selected methods in the image;
The image processing apparatus according to claim 1, wherein the method identification information embedding unit embeds the method identification information by a method that embeds less information than any of the methods used by the information embedding unit.   12. The image processing apparatus according to claim 11, wherein the system identification information embedding unit synthesizes a predetermined pattern with the background of the image.   13. The image according to claim 12, wherein the scheme identification information embedding unit synthesizes the predetermined pattern and a pattern obtained by rotating the predetermined pattern in accordance with a value of the scheme identification information with the image. Processing equipment.   The image processing apparatus according to claim 12, wherein the predetermined pattern is a pattern composed of a plurality of dots. An image processing apparatus that extracts information from an image in which information may be embedded by a plurality of methods,
Method identification information extracting means for extracting method identification information for identifying a method used for embedding information from the image;
Information extracting means for extracting information embedded in the image using one or more methods identified by the method identification information;
The image processing apparatus according to claim 1, wherein the method identification information is embedded in the image by a method that embeds less information than any method used by the information extraction unit. First storage means capable of storing a part of the image;
A second storage means capable of storing all of the image;
Image reading means for reading the image from a medium;
Storage means for storing the read image in the first storage means and the second storage means, wherein the method identification information extraction means is the above-described image stored in the first storage means Extract method identification information,
The image processing apparatus according to claim 15, wherein the information extraction unit extracts the information from an image stored in the second storage unit.   The image processing apparatus according to claim 15, wherein the information extraction unit does not perform extraction of the information according to a value of the method identification information.   18. The image processing apparatus according to claim 17, wherein the method identification information extracting unit extracts the method identification information based on a predetermined pattern synthesized with a background of the image.   19. The method identification information extraction unit according to claim 18, wherein the method identification information extraction unit determines a value of the method identification information based on an angle difference between the predetermined pattern and a pattern obtained by rotating the predetermined pattern. Image processing device.   The image processing apparatus according to claim 18, wherein the predetermined pattern is a pattern composed of a plurality of dots.   An image processing program for causing a computer to execute the image processing method according to claim 1.   An image processing program for causing a computer to execute the image processing method according to claim 5.

Images