JP4192906B2 - Watermark information detection apparatus and watermark information detection method - Google Patents

Description

  The present invention relates to a technique for detecting confidential information from a printed document including confidential information or a document image including confidential information.

  “Digital watermark”, which embeds information to prevent copying / counterfeiting or confidential information (or confidential information) in an image or document data in a form that is invisible to the human eye, is all stored and passed on electronic media. Therefore, the information embedded in the watermark (watermark information) is not deteriorated or lost, so that the information can be reliably detected.

  As with the above electronic media, documents printed on paper media can be easily tampered with in a form that is not visually unsightly other than text to prevent unauthorized tampering and copying. This is possible by a method of embedding confidential information in a printed document.

The following methods (1) to (7) are known as methods for detecting confidential information embedded in black and white binary documents that are most widely used as printed materials (for example, Patent Documents). 1).
(1) A printed document with a watermark embedded is scanned to obtain a document image.
(2) A filtering process is performed to obtain an output value from each filter that reacts to each signal unit embedded in the entire document image as watermark information.
(3) The peak value of the filter that reacted strongly at the interval between each signal unit is searched. The search is performed centering on a location separated from the adjacent signal unit position by the interval of the signal unit used at the time of embedding. The search is performed on the entire filtering process result.
(4) A signal unit is identified by comparing the magnitudes of peak values output from the filters, and information associated with the signal unit (for example, a bit value of “0” or “1”) is extracted. To do.
(5) The effective range in which the information is embedded is specified from the extracted information.
(6) The information is combined in the format used when the extracted information is embedded in the print document, and the entire watermark information is acquired.
(7) It is possible to detect the confidential information (bit string) by correcting the erroneous information by performing a majority operation from the entire watermark information and acquiring the correct information.

  Here, with reference to FIG. 1, (2) the filtering process that reacts to each signal unit will be described. FIG. 1 is an explanatory diagram illustrating an example of an outline of filtering processing.

  As shown in FIG. 1, a document image 17 obtained by scanning a printed document has a stain due to, for example, ink. The document image 17a is an enlarged part of the dirt.

  The document image 17 a is composed of a plurality of blocks 18. One block 18 represents one pixel.

  Next, when filtering the document image 17a, the Gabor filter A and the Gabor filter B are used. The Gabor filters A and B are both composed of 12 × 12 pixels. It is not limited.

  When the filtering process is performed on the document image 17a by the Gabor filters A and B, the document image 17b and the document image 17c are generated by obtaining the output values that are reacted by the Gabor filters.

  Further, the document image 17b shown in FIG. 1 is obtained by superimposing the document image 17b and the document image 17c.

  As shown in FIG. 1, when there is dirt on the document image, the signal unit cannot be identified and erroneous information may be detected because the reaction of the dirty Gabor filter is weak. there were.

  Therefore, in order to solve the above problem, the confidential information embedded in the entire document image is provided with redundancy such that the same information is repeatedly embedded several times, and the majority information is calculated in the above (7), so that incorrect information is obtained. It can be corrected.

JP 2003-101762 A

  However, when the degradation on the document image is severe, many of the extracted signal units have low reliability, and if they are used as they are for the majority operation, only information with high reliability suitable for decoding can be used. However, as shown in FIG. 2, there is a problem that incorrect information is included in the confidential information finally obtained by majority calculation.

  In addition to the dirt on the document image, for example, even if the signal processing in which a character or a figure overlaps, the filtering process shown in the above (2) reacts strongly to the edge component of the character or figure. It was not an appropriate reaction, and there was a problem that incorrect information was included in the confidential information that was finally obtained by majority calculation.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide a new and improved watermark information detection apparatus and watermark information detection method capable of improving the accuracy of decryption to confidential information. Is to provide.

  In order to solve the above problems, according to a first aspect of the present invention, there is provided a watermark information detecting apparatus for extracting confidential information from an image. The watermark information detection apparatus includes an image input unit that inputs an image and obtains the input image; and extracts one or more patterns embedded in the input image by filtering the input image and supports the pattern A pattern determining unit that determines and determines a symbol to be determined; a probability calculating unit that calculates the accuracy of determination of the symbol for each pattern embedded in the input image; and a pattern selection based on the calculated determination accuracy And an information decrypting unit for decrypting the selected pattern into confidential information, and the information decrypting unit further executes a certainty factor calculation on the selected pattern and decrypts it into the confidential information. It is said.

  According to the present invention, the watermark information detection apparatus acquires one or more patterns embedded in an input image by filtering, and determines a corresponding symbol. Further, the watermark information detection apparatus obtains the accuracy of determining that the symbol is determined, performs pattern selection based on the accuracy of the determination, selects only a pattern having a high accuracy of determination, and uses the selected pattern for confidential information. Is decrypted. This configuration improves the accuracy of decryption to confidential information because it can be decrypted into confidential information after eliminating patterns with low accuracy in determining the symbol corresponding to the pattern from the input image due to character pixels and dirt. Can be made.

In the certainty factor calculation, a majority vote calculation for taking a majority vote for the selected pattern may be performed.
In the certainty calculation, the accumulated values of a plurality of filter output values obtained by accumulating the filter output values for each filter may be compared with the selected pattern. In the certainty calculation, You may compare the magnitude | size of the accumulated value of several determination accuracy calculated | required by accumulating determination accuracy for every filter with respect to the said selected pattern.
Further, the certainty factor calculation may be configured to perform a majority operation on patterns that are repeatedly embedded in the input image and that have the same repetitive embedding order.
In addition, the certainty calculation is performed by accumulating a plurality of filter output values obtained by accumulating filter output values for each filter with respect to patterns having the same repetitive embedding order among the patterns repeatedly embedded in the input image. The above certainty factor calculation is obtained by accumulating the determination accuracy for each filter for the patterns that are repeatedly embedded in the input image and having the same repetition order. You may comprise so that it may be the calculation which compares the magnitude of the cumulative value of several determination accuracy.

In order to solve the above problems, according to another aspect of the present invention, one or more identifiable patterns are prepared, at least one symbol is given to one pattern, and confidential information is encoded. An image input unit that reads an image in which the confidential information is embedded as an input image by arranging one or two or more of the patterns generated as described above; Filter the input image by preparing the same type of detection filter that reacts to each of two or more patterns, and based on the output values obtained by all types of detection filters in the image area of the input image A pattern determining unit for determining and determining a symbol of the pattern; based on the output value for each pattern obtained by the above-described all types of detection filters; An accuracy calculation unit for calculating the accuracy of symbol determination for each pattern embedded in the input image; and selecting a pattern corresponding to the determination accuracy according to the determination accuracy calculated by the determination accuracy calculation unit. An information decrypting unit that decrypts the confidential information with the selected pattern, and the information decrypting unit further executes a certainty factor operation on the selected pattern to decrypt the confidential information. It is characterized by. Examples of the pattern according to the present invention include a signal unit and / or a unit pattern obtained by unitizing a plurality of patterns.

  The determination accuracy may be configured to be the maximum value among the output values obtained by filtering with all types of detection filters, and the determination accuracy is obtained by filtering with all types of detection filters. The output value may be a difference value.

The input image may be configured such that a unit pattern having at least one or more patterns configured by encoding confidential information as a unit is repeatedly embedded several times. With this configuration, it is possible to improve the accuracy of decryption to confidential information.

The certainty factor calculation may be an operation for performing majority calculation for unit patterns having the same order of repeated embedding among unit patterns repeatedly embedded in the input image. Or you may comprise so that it may be a calculation which performs majority calculation about the unit patterns in which the number of bit digits of the confidential information obtained by decoding one or two or more unit patterns corresponds to the same number of bits. With this configuration, it is possible to improve the accuracy of decryption to confidential information after correcting to a correct pattern.

  The information decoding unit may be configured to calculate a threshold value of the determination accuracy, compare the threshold value of the determination accuracy with the determination accuracy, and determine the selection of the pattern according to the comparison result. With this configuration, it is possible to reduce the possibility that an erroneous pattern exists in the pattern used in the majority operation, and to decrypt the confidential information with higher accuracy.

  The determination accuracy threshold may be calculated and held in advance before selection by the information decoding unit, and the determination accuracy threshold may be one or more calculated determinations. You may comprise so that it may be an average value of accuracy.

The watermark information detecting device further comprises an error detection unit, the error detection unit may be configured to detect an error by the error detection code to the secure information decoded by the information decoding section. With such a configuration, the integrity of confidential information can be further improved and the accuracy of confidential information can be improved as compared with the case where only the majority operation is performed.

  The error detection unit performs error detection on the pattern selected by the information decoding unit, and as a result, when the error is detected, the information decoding unit selects a pattern from the plurality of determination accuracy. A configuration may be adopted in which pattern selection is performed again based on a determination accuracy different from the determination accuracy used, a maximum likelihood determination is further performed on the selected pattern, and the confidential information is decrypted. Good. With this configuration, even if the determination of the symbol corresponding to the pattern depends on the printing environment or the like, the accuracy of decryption to confidential information can be improved under any environment by changing the threshold value of the appropriate determination accuracy.

  The information decoding unit may be configured to repeatedly execute maximum likelihood determination after error detection by the error detection unit until no error is detected by the error detection unit. With this configuration, it is possible to improve the accuracy of decryption to confidential information to the maximum extent.

  The information decoding unit may be configured to use a plurality of determination accuracy thresholds one by one in order. With this configuration, it is possible to determine an appropriate determination accuracy threshold.

The information decoding unit, the threshold value of the determination accuracy may be configured to dynamically calculate differently. With such a configuration, it is possible to determine an appropriate determination accuracy threshold value under any environment such as a printing environment.

  According to another aspect of the present invention for solving the above problem, a watermark information detection method for extracting confidential information from an image is provided. The watermark information detection method includes an image input step of inputting an image and obtaining the input image; filtering the input image to extract one or more patterns embedded in the input image and corresponding to the pattern A pattern determining step for determining and determining a symbol to be determined; a probability calculating step for calculating the accuracy of symbol determination for each pattern embedded in the input image; and selecting a pattern based on the calculated determination accuracy; And an information decrypting step for decrypting the selected pattern into confidential information. In the information decrypting step, a maximum likelihood determination is further performed on the selected pattern to decrypt it into the confidential information.

  According to another aspect of the present invention for solving the above problem, one or more identifiable patterns are prepared, at least one symbol is given to one pattern, and confidential information is encoded. An image input process that reads an image in which the confidential information is embedded as an input image by arranging one or more patterns generated by the conversion; and reacts to the pattern to detect the pattern from the input image Prepare only the same type of detection filter as the pattern to filter the input image, and determine and determine the symbol of the pattern based on the output values obtained by all types of detection filters in the image area of the input image The pattern determination step; and the accuracy of symbol determination is improved based on the output value of each pattern obtained by all types of detection filters. An accuracy calculation step for calculating each pattern embedded in the input image; according to the determination accuracy calculated in the determination accuracy calculation step, a pattern corresponding to the determination accuracy is selected, and the selected pattern is classified as confidential. An information decrypting step for decrypting the information. In the information decrypting step, a maximum likelihood determination is further performed on the selected pattern to decrypt it into confidential information.

  As described above, according to the present invention, it is possible to improve the accuracy of decrypting confidential information. Further, it is possible to improve the accuracy of decrypting the confidential information without being influenced by an environment such as a printing environment.

  DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, preferred embodiments of the invention will be described in detail with reference to the accompanying drawings. In the following description and the accompanying drawings, components having substantially the same functions and configurations are denoted by the same reference numerals, and redundant description is omitted.

(First embodiment)
FIG. 3 is an explanatory diagram showing the configuration of the watermark information embedding device and the watermark information detecting device according to this embodiment.

(Watermark information embedding device 10)
The watermark information embedding device 10 is a device that forms a document image based on document data and confidential information embedded in the document and prints it on a paper medium. As shown in FIG. 3, the watermark information embedding device 10 includes a document image forming unit 11, a watermark image forming unit 12, a watermarked document image synthesizing unit 13, and an output device 14. The document data 15 is data created by a document creation tool or the like. The confidential information 16 is information (character string, image, audio data) to be embedded in a paper medium in a format other than characters.

  In the document image forming unit 11, an image in a state where the document data 15 is printed on a paper surface is created. Specifically, the white pixel region in the document image is a portion where nothing is printed, and the black pixel region is a portion where black paint is applied. In this embodiment, the description will be made on the assumption that printing is performed on white paper with black ink (single color). However, the present invention is not limited to this, and printing is performed in color (multicolor). However, the present invention can be similarly applied.

  The watermark image forming unit 12 digitizes the confidential information 16 and converts it into a numerical value, performs N-ary coding (N is 2 or more), and assigns each symbol of the code word to a signal prepared in advance. A signal represents a wave having an arbitrary direction and wavelength by arranging dots in a rectangular area of an arbitrary size, and a symbol is assigned to the direction and wavelength of the wave. A watermark image is one in which these signals are arranged on the image according to a certain rule.

  The watermarked document image synthesis unit 13 creates a watermarked document image (for example, an input image) by superimposing the document image and the watermark image. The output device 14 is an output device such as a printer, and prints a watermarked document image on a paper medium. Therefore, the document image forming unit 11, the watermark image forming unit 12, and the watermarked document image synthesizing unit 13 may be realized as one function in the printer driver.

  The print document 20 is printed with the confidential information 16 embedded in the original document data 15 and is physically stored and managed. The print document 20 according to the present embodiment will be described by taking a paper medium or the like as an example. However, the print document 20 is not limited to this example, and can be implemented, for example, even when image data including a document image is used as it is. is there.

(Watermark information detection device 30)
The watermark information detecting device 30 is a device that takes in a document printed on a paper medium (for example, a printed document) as an input image and restores embedded confidential information. As shown in FIG. 3, the watermark information detection apparatus 30 includes an input device (image input unit) 31 and a watermark detection unit 32.

  The input device 31 is an input device such as a scanner, and takes in the document 20 printed on paper as a multi-value gray image into a computer. Also, the watermark detection unit (pattern determination unit, accuracy calculation unit, information decoding unit) 32 performs a filtering process on the input image and detects an embedded signal. The symbol is restored from the detected signal, and the embedded confidential information is extracted. The watermark detection unit according to the present embodiment corresponds to at least one of the pattern determination unit, the accuracy calculation unit, and the information decoding unit according to the present invention, but is not limited to this example.

  Operations of the watermark information embedding device 10 and the watermark information detection device 30 configured as described above will be described. First, the operation of the watermark information embedding device 10 will be described with reference to FIGS.

(Document Image Forming Unit 11)
The document data 15 is data including font information and layout information, and is created by word processing software or the like. Based on the document data 15, the document image forming unit 11 creates an image of a document printed on paper for each page. This document image is a black and white binary image. On the image, white pixels (pixels having a value of 1) are backgrounds, and black pixels (pixels having a value of 0) are character regions (regions to which ink is applied). It shall be.

(Watermark image forming unit 12)
The confidential information 16 is various data such as characters, sounds, and images, and the watermark image forming unit creates a watermark image to be superimposed as the background of the document image from this information.

  FIG. 4 is a flowchart showing a processing flow of the watermark image forming unit 12.

  First, the confidential information 16 is converted into an N-element code (step S101). N is arbitrary, but in the present embodiment, N = 2 for ease of explanation. Accordingly, the code generated in step S101 is a binary code, and is expressed by a bit string of 0 and 1. In this step S101, the data may be encoded as it is, or the encrypted data may be encoded.

  Next, a watermark signal is assigned to each symbol of the codeword (step S102). The watermark signal represents a wave having an arbitrary wavelength and direction by an arrangement (dot pattern) of dots (black pixels). The watermark signal will be further described later.

  Further, a signal unit corresponding to the bit string of the encoded data is arranged on the watermark image (step S103).

  The watermark signal assigned to each symbol of the code word in step S102 will be described. FIG. 5 is an explanatory diagram showing an example of a watermark signal.

  Let the width and height of the watermark signal be Sw and Sh, respectively. Sw and Sh may be different, but in the present embodiment, Sw = Sh is set for ease of explanation. The unit of length is the number of pixels. In the example of FIG. 5, Sw = Sh = 12. The size when these signals are printed on the paper surface depends on the resolution of the watermark image. For example, the watermark image is an image of 600 dpi (dot per inch: unit of resolution, number of dots per inch). 5, the width and height of the watermark signal in FIG. 5 is 12/600 = 0.02 (inch) on the printed document.

  Hereinafter, a rectangle having a width and a height of Sw and Sh is referred to as a “signal unit” as one signal unit. In FIG. 5A, the distance between dots is dense in the direction of arctan (3) (arctan is an inverse function of tan) with respect to the horizontal axis, and the wave propagation direction is arctan (−1/3). . Hereinafter, this signal unit is referred to as unit A. In FIG. 5B, the distance between dots is dense in the direction of arctan (−3) with respect to the horizontal axis, and the wave propagation direction is arctan (1/3). Hereinafter, this signal unit is referred to as unit B.

6, arctan a change in pixel value in FIG. 5 (1) - is a sectional view seen from the direction of the (1/3). In FIG. 6, the portion where the dots are arranged is the antinode of the minimum value of the wave (the point where the amplitude is maximum), and the portion where the dots are not arranged is the antinode of the maximum value of the wave.

  In addition, since there are two regions in each unit where dots are densely arranged, the frequency per unit is 2 in this example. Since the wave propagation direction is perpendicular to the direction in which the dots are densely arranged, the wave of unit A is arctan (-1/3) with respect to the horizontal direction, and the wave of unit B is arctan (1/3). Become. Note that a × b = −1 when the direction of arctan (a) and the direction of actan (b) are perpendicular.

  In the present embodiment, symbol 0 is assigned to the watermark signal expressed by unit A, and symbol 1 is assigned to the watermark signal expressed by unit B. These are called symbol units.

  In addition to the watermark signals shown in FIGS. 5 (1) and (2), for example, dot arrangements as shown in FIGS. 7 (3) to (5) are conceivable. In FIG. 7 (3), the distance between dots is dense in the direction of arctan (1/3) with respect to the horizontal axis, and the wave propagation direction is arctan (-3). Hereinafter, this signal unit is referred to as unit C.

  In FIG. 7 (4), the distance between dots is dense in the direction of arctan (−1/3) with respect to the horizontal axis, and the wave propagation direction is arctan (3). Hereinafter, this signal unit is referred to as unit D. In FIG. 7 (5), the distance between the dots is dense in the direction of arctan (1) with respect to the horizontal axis, and the wave propagation direction is arctan (−1). In FIG. 7 (5), it can be considered that the distance between the dots is dense in the direction of arctan (−1) with respect to the horizontal axis, and the wave propagation direction is arctan (1). Hereinafter, this signal unit is referred to as unit E.

  In this way, in addition to the combinations assigned previously, there can be a plurality of combinations of units to which symbols 0 and 1 are assigned. Therefore, it is possible to keep secret which watermark signal is assigned to which symbol. It is also possible to prevent a person (illegal person) from easily deciphering the embedded signal.

  Furthermore, when the confidential information is encoded with a quaternary code in step S102 shown in FIG. 4, for example, the code word symbol 0 is assigned to unit A, symbol 1 is assigned to unit B, and symbol 2 is assigned to unit C. , It is also possible to assign symbol 3 to unit D

  In the example of the watermark signal shown in FIGS. 5 and 7, since the number of dots in one unit is all equal, by arranging these units without gaps, the apparent density of the watermark image becomes uniform. . Therefore, it appears that a gray image having a single density is embedded as a background on the printed paper.

  In order to produce such an effect, for example, unit E is defined as a background unit (signal unit to which no symbol is assigned), and this is arranged without any gap as a background of the watermark image, and symbol unit (unit A, unit B) Is embedded in the watermark image, the background unit (unit E) and the symbol unit (unit A, unit B) at the position to be embedded are exchanged.

  FIG. 8A is an explanatory diagram showing a case where the unit E is defined as a background unit and arranged as a background of a watermark image without gaps. FIG. 8 (2) shows an example in which the unit A is embedded in the background image of FIG. 8 (1), and FIG. 8 (3) shows an example in which the unit B is embedded in the background image of FIG. 8 (1). Show. In the present embodiment, a method of using a background unit as the background of a watermark image will be described. However, a watermark image may be generated by arranging only symbol units.

  Next, a method for embedding one symbol of a code word in a watermark image will be described with reference to FIG.

  FIG. 9 is an explanatory diagram illustrating an example of a symbol embedding method in a watermark image. Here, a case where a bit string “0101” is embedded will be described as an example.

  As shown in FIGS. 9A and 9B, the same symbol unit is repeatedly embedded. This is to prevent the character unit in the document from being detected when a signal is detected when it is superimposed on the embedded symbol unit. The symbol unit repetition number and arrangement pattern (hereinafter referred to as a unit pattern) are used. Is optional.

  That is, as an example of the unit pattern, the repetition number is set to 4 (4 symbol units exist in one unit pattern) as shown in FIG. 9 (1), or the repetition number is set to 2 as shown in FIG. 9 (2). (There are two symbol units in one unit pattern) or the number of repetitions may be one (only one symbol unit exists in one unit pattern).

  9 (1) and 9 (2), one symbol is given to one symbol unit. However, as shown in FIG. 9 (3), symbols may be given to the symbol unit arrangement pattern. good.

  The number of bits of information that can be embedded in a watermark image for one page depends on the size of the signal unit, the size of the unit pattern, and the size of the document image. The number of signals embedded in the horizontal and vertical directions of the document image may be detected as a known signal, or may be calculated backward from the size of the image input from the input device and the size of the signal unit.

  If Pw unit patterns in the horizontal direction and Ph units in the vertical direction can be embedded in a watermark image for one page, unit patterns at arbitrary positions in the image are represented by U (x, y), x = 1 to Pw, y = 1 to Ph, and U (x, y) is referred to as a “unit pattern matrix”. In addition, the number of bits that can be embedded in one page is referred to as the “number of embedded bits”. The number of embedded bits is Pw × Ph.

  FIG. 10 is a flowchart showing a method for embedding confidential information in a watermark image.

  Here, a case where the same information is repeatedly embedded in one (one page) watermark image will be described. By repeatedly embedding the same information, it is possible to extract the embedded information even if the embedded information disappears due to, for example, the entire unit pattern being filled when the watermark image and the document image are overlaid. Because.

  First, the confidential information 16 is converted into an N-element code (step S201). This is the same as step S101 in FIG. Hereinafter, the encoded data is referred to as a data code, and the data code expressed by a combination of unit patterns is referred to as a data code unit Du.

  Next, based on the code length of the data code (here, the number of bits) and the number of embedded bits, how many times the data code unit can be embedded in one image is calculated (step S202). In the present embodiment, it is assumed that the code length data of the data code is inserted into the first row of the unit pattern matrix. The code length of the data code may be fixed and the code length data may not be embedded in the watermark image.

  The number Dn of embedding the data code unit is calculated by the following equation with the data code length as Cn.

Here, assuming that the remainder is Rn (Rn = Pw × (Ph−1) −Dn × Cn ), the unit pattern matrix is embedded with Dn data code units and a unit pattern corresponding to the first Rn bits of the data code. become. However, the Rn bit of the remainder part does not necessarily have to be embedded.

  In the description of FIG. 11, the size of the unit pattern matrix is 9 × 11 (11 rows and 9 columns), and the data code length is 12 (numbers 0 to 11 in the figure indicate each code word of the data code). And

  Next, the code length data is embedded in the first row of the unit pattern matrix (step S203). In the example of FIG. 11, the code length is expressed by 9-bit data and is embedded only once. However, if the unit pattern matrix width Pw is sufficiently large, It can be embedded repeatedly.

  Further, the data code unit is repeatedly embedded in the second and subsequent rows of the unit pattern matrix (step S204). As shown in FIG. 11, the MSB (most significant bit) or LSB (least significant bit) of the data code is sequentially embedded in the row direction. The example of FIG. 11 shows an example in which the data code unit is embedded seven times and the first 6 bits of the data code are embedded.

  The data embedding method may be embedded so as to be continuous in the row direction as shown in FIG. 11 or may be embedded so as to be continuous in the column direction.

  The watermark image in the watermark image forming unit 12 has been described above. Next, the watermarked document image composition unit 13 of the watermark information embedding device 10 will be described.

(Watermarked document image composition unit 13)
The watermarked document image composition unit 13 superimposes the document image created by the document image forming unit 11 and the watermark image created by the watermark image forming unit. The value of each pixel of the watermarked document image is calculated by a logical product operation (AND) of the corresponding pixel values of the document image and the watermark image. That is, if either the document image or the watermark image is 0 (black), the pixel value of the watermarked document image is 0 (black), otherwise 1 (white).

  FIG. 12 is an explanatory diagram showing an example of a watermarked document image. FIG. 13 is an explanatory view showing a part of FIG. 12 in an enlarged manner. Here, the unit pattern shown in FIG. 9A is used. The watermarked document image is output by the output device 14.

  The operation of the watermark information embedding device 10 has been described above.

  Next, the operation of the watermark information detection apparatus 30 will be described with reference to FIGS. 3 and 14 to 22.

(Watermark detection unit 32)
FIG. 14 is a flowchart showing a process flow of the watermark detection unit 32.
First, a watermarked document image is input to a memory of a computer or the like by an input device 31 such as a scanner (step S301). This image is referred to as an input image. The input image is a multi-valued image, and will be described below as a gray image with 256 gradations. Further, the resolution of the input image (resolution when read by the input device 31) may be different from the watermarked document image created by the watermark information embedding device 10, but here the image created by the watermark information embedding device 10 is used. It is assumed that the resolution is the same. Also, it is assumed that the input image has been corrected for rotation, expansion and contraction.

Next, the number of unit patterns embedded is calculated from the size of the input image and the size of the signal unit (step S302). For example, assuming that the size of the input image is W (width) × H (height), the size of the signal unit is Sw × Sh, and the unit pattern is composed of Uw × Uh signal units. The number of unit patterns embedded therein (N = Pw × Ph) is calculated as follows.

  However, when the resolution is different between the watermark information embedding device 10 and the watermark information detection device 30, the above calculation is performed after normalizing the size of the signal unit in the input image based on the ratio of the resolutions.

  Next, a unit pattern separation position is set for the input image based on the number of unit patterns calculated in step S302 (step S303). FIG. 15 shows an example of the input image (FIG. 15 (1)) and the input image (FIG. 15 (2)) after setting the unit pattern separation position.

  Next, a symbol unit is detected for each unit pattern delimiter, and a unit pattern matrix is restored (step S304). Details of signal detection will be described below.

  FIG. 16 is an explanatory diagram showing an example of an area corresponding to the unit A shown in FIG. 5A in the input image. In FIG. 5, the signal unit is a binary image, but here it is a multi-valued image. When a binary image is printed, the density changes continuously due to ink bleeding and the like, resulting in a white and black gradation, so that the periphery of the dot is an intermediate color between white and black as shown in FIG. Therefore, a cross-sectional view of FIG. 16 viewed from a direction parallel to the wave propagation direction is as shown in FIG. 6 is a rectangular wave, whereas FIG. 17 is a smooth wave.

  Also, in reality, many noise components are added to the input image due to factors such as local changes in paper thickness, dirt on the printed document, and instability of the output device and image input device. However, the case where there is no noise component will be described here. However, if the method described here is used, stable signal detection can be performed even for an image to which a noise component is added.

  In the following, in order to detect the signal unit from the input image, a two-dimensional wavelet filter (detection filter) capable of simultaneously defining the frequency and direction of the wave and the influence range is used. In the following, an example using a Gabor filter, which is one of the two-dimensional wavelet filters, is shown. However, if it is a filter having the same properties as the Gabor filter, it is not necessarily a Gabor filter, and moreover, the same dot as the signal unit. A method of defining a template having a pattern and performing pattern matching may be used.

  The Gabor filter G (x, y), x = 0 to gw−1, and y = 0 to gh−1 are shown below. gw and gh are filter sizes, which are the same size as the signal unit embedded by the watermark information embedding apparatus 10 here.

For signal detection, the same number of Gabor filters as the types of embedded signal units are prepared with the same frequency, wave direction, and size as the signal units embedded in the watermark image. Here, the Gabor filters corresponding to the units A and B in FIG.

  The filter output value at an arbitrary position in the input image is calculated by convolution between the filter and the image. In the case of a Gabor filter, there are a real number filter and an imaginary number filter (an imaginary number filter is a filter whose phase is shifted by a half wavelength from the real number filter), and the mean square value thereof is used as a filter output value. For example, if the convolution between the real filter and the image of the filter A is Rc and the convolution between the imaginary filter is Ic, the output value F (A) is calculated by the following equation.

  FIG. 18 is an explanatory diagram for explaining a method for determining whether the symbol unit embedded in the unit pattern U (x, y) delimited in step S303 is the unit A or the unit B.

The symbol determination step for the unit pattern U (x, y) is performed as follows.
(1) While moving the position of the filter A, the maximum value of the result of calculating F (A) for all the positions in the unit pattern U (x, y) is the value of the filter A for the unit pattern U (x, y). The output value is set to Fu (A, x, y).
(2) The output value of the filter B for the unit pattern U (x, y) is calculated in the same manner as (1), and this is assumed to be Fu (B, x, y).
(3) Fu (A, x, y) and Fu (B, x, y) are compared, and if Fu (A, x, y) ≧ Fu (B, x, y), unit pattern U (x, y) The symbol unit embedded in y) is determined to be unit A, and if Fu (A, x, y) <Fu (B, x, y), it is embedded in the unit pattern U (x, y). It is determined that the existing symbol unit is unit B.
(4) The symbol unit determination accuracy (unit determination accuracy) determined to be embedded in the unit pattern U (x, y ) is P (x, y), and the maximum output value of the filters A and B Value or the difference value between the output values of filters A and B. The formula for calculating the unit determination accuracy P (x, y) in the case of the maximum value of the output value of the filter or the difference value of the output value of the filter is shown below. The unit determination accuracy may be simply referred to as determination accuracy.

  In the above (1) and (2), the step width for moving the filter is arbitrary, and only the output value at a representative position on the unit pattern may be calculated. If the absolute value of the difference between Fu (A, x, y) and Fu (B, x, y) is less than or equal to a predetermined threshold value in (3), “P (x, y) = 0 ”, and the determination may be impossible.

  The details of the signal detection (step S304) have been described above. Returning to the flowchart of FIG. 14 again, the following step S305 will be described. In step S305, the symbols of the unit pattern matrix are concatenated to reconstruct the data code, and the original information is restored.

FIG. 19 is an explanatory diagram showing an example of information restoration. The steps of information restoration are as follows.
(1) A symbol embedded in each unit pattern is detected ((1) shown in FIG. 19).
(2) The symbols are connected to restore the data code ((2) shown in FIG. 19).
(3) Decode the data code to extract the embedded information ((3) shown in FIG. 19).

  20 to 22 are explanatory diagrams illustrating an example of a data code restoration method. The restoration method is basically the reverse process of FIG.

  First, the code length data portion is extracted from the first row of the unit pattern matrix to obtain the code length of the embedded data code (step S401).

  Next, based on the size of the unit pattern matrix and the code length of the data code obtained in S401, the number Dn of data code units embedded and the remainder Rn are calculated (step S402).

Next, data code units are extracted from the second and subsequent rows of the unit pattern matrix by the reverse method of step S203 (step S403). In the example of FIG. 21, each unit pattern is decomposed in order from U (2, 1) (2 rows and 1 column). For example, the first unit pattern is U (2, 1) to U (3, 3), and the second unit pattern is U (3, 4) to U (4, 6). is there. The same applies to the third unit pattern and thereafter. Since Dn = 7 and Rn = 6, 12 unit patterns (data code units) are extracted 7 times, and 6 unit patterns (corresponding to the upper 6 data code units) (U (11 , 4) to U (11, 9)) are taken out. Further, the unit determination accuracy P (x, y) already obtained is similar to the case where U (x, y) is decomposed into 12 unit patterns sequentially from U (2, 1) (2 rows and 1 column). ) Is also decomposed so as to correspond to the above U (x, y) every 12 unit patterns in order from P (2, 1). For example, among the unit determination accuracy P (x, y), the unit determination accuracy associated with the first data code unit is P (2,1) to P (3,3), and the second data code unit. The unit determination accuracy associated with is such as P (3, 4) to P (4, 6).

  Next, the embedded data code is reconstructed by performing bit certainty calculation on the data code unit and unit determination accuracy extracted in step S403 (step S404). Hereinafter, the bit certainty calculation will be described.

  As shown in FIG. 22A, the data code units first extracted from the second row and the first column of the unit pattern matrix are Du (1,1) to Du (1,12), and Du (2,1) to Du (2 , 12),... Further, the surplus portion is assumed to be Du (8, 1) to Du (8, 6). As shown in FIG. 22B, the unit determination accuracy already obtained in the above is also a series of unit patterns (P (2, 1) to P (3, 3), P ( (3, 4) to P (4, 6),..., Pu (1, 1) to Pu (1, 12), Pu (2, 1) to Pu (2, 12),. Further, the surplus portion is assumed to be Pu (8, 1) to Pu (8, 6). The bit certainty calculation is a process of determining the value of each symbol of the data code by taking a majority vote for each element of each data code unit. At this time, by selectively using an effective data code unit based on the unit determination accuracy, signals from any unit in any data code unit can be correctly signaled due to overlap with the character area or dirt on the paper. Even if the detection cannot be performed (such as a bit inversion error), the data code can be finally restored correctly.

  Here, with reference to FIG. 22C, the processing for selecting an effective data code unit based on the unit determination accuracy will be described. As shown in FIG. 22C, the unit determination accuracy Pu that is equal to or greater than the threshold T is extracted. To do. As shown in FIG. 22C, Pu (x, y) marked with a circle indicates that the unit determination accuracy Pu of Pu (x, y) is smaller than the threshold T.

  For example, as shown in FIG. 22C, if Pu (x, y) smaller than the threshold T is selected from Pu (4, 1) to Pu (4, 12), Pu (4, 3), Pu (4, 5), Pu (4, 6), Pu (4, 7), Pu (4, 9), Pu (4, 12).

  When all Pu (x, y) smaller than the threshold T are selected, all Du (x, y) corresponding to Pu (x, y) smaller than the threshold T are selected and excluded (erased). (Arrow (1) shown in FIG. 22C). That is, Du (x, y) corresponding to Pu (x, y) smaller than the threshold T has low reliability, and the symbol itself is highly likely to be erroneous, and the Du (x, y) is excluded. By doing so, it is possible to improve the accuracy of error correction in the majority operation.

In the state where Du (x, y) related to Pu (x, y) smaller than the threshold T is excluded, the majority operation is executed as shown in FIG. 22D. In particular, for example, for the array [5] (Du (1, 6), Du (2, 6),..., Du (8, 6)), the one having more 1 or 0 is selected from the three by a majority decision. Store in the reconstructed bit string shown in FIG. 22D. In the conventional bit string (array [5]), there are already five erroneous values even after the majority operation, so the possibility of correcting the error and storing the correct value in the bit string can be improved. There wasn't.

  More specifically, the majority operation will be described. For example, the first bit of the data code is a signal of Du (1, 1), Du (2, 1), Du (3, 1), ..., Du (8, 1). Of the detection results, the majority of the signal detection results with unit determination accuracy Pu (Pu (1,1), Pu (2,1), Pu (3,1),..., Pu (8,1)) equal to or greater than the threshold value T are determined. When there are more “1” s in the calculation, it is determined as 1 and when there are more “0” s, it is determined as 0. The threshold T of the unit determination accuracy Pu can be specifically expressed by the following equation.

  Similarly to the above, the second bit of the data code includes the signal detection results of Du (1, 2), Du (2, 2), Du (3, 2),. The accuracy Pu (Pu (1, 2), Pu (2, 2), Pu (3, 2),..., Pu (8, 2)) is determined based on the signal detection result equal to or greater than the threshold T, and so on. Then, the determination is made up to the 12th bit.

  The bit certainty calculation can also be performed by adding the output values of the signal detection filter of FIG. That is, it is possible to calculate the bit certainty factor by obtaining the cumulative value of the filter output value for each filter and comparing the cumulative values of the obtained plural filter output values. This is because, for example, a symbol of 0 is assigned to unit A in FIG. 5 (1), and a symbol of 1 is assigned to unit B in FIG. 5 (2). If the maximum value of the output value is Df (A, m, n), and the maximum value of the output value by the filter B for Du (m, n) is Df (B, m, n), N of the data code serving as confidential information The bit number is obtained by the following conditional expression.

When the relationship of the above formula is established, it is determined as 0, and otherwise it is determined as 1 . However, when N <Rn, the addition of Df is n = 1 to Rn + 1. For example, for the signal detection result of the second bit of the data code (Du (1, 2), Du (2, 2), Du (3, 2),..., Du (8, 2)), the output of the signal detection filter If the values are accumulated and the relationship of the above equation is established, it is determined as 0 . The bit certainty calculation is not limited to the case where it is executed based on the accumulated value of the output value of the signal detection filter. It is also possible to execute the bit certainty calculation by comparing the accumulated values of a plurality of unit determination accuracy (or determination accuracy) obtained by accumulating the unit determination accuracy for each filter in the equation.

  Although the case where data codes are repeatedly embedded has been described here, a method that does not repeat data code units can be realized by using an error correction code or the like when encoding data.

  Further, after executing the majority decision of the bit certainty calculation, an error detection code or the like is used for the decoded confidential information bit string (data code) as shown in FIG. Even if it is a case, it can be implemented.

As described above in detail, according to the present embodiment, the following excellent effects are obtained.
(1-1) Since the embedded information is expressed by the difference in dot arrangement, the original document font, character spacing, and line spacing are not changed.
(1-2) Since the dot pattern to which symbols are assigned and the dot pattern to which symbols are not assigned have the same density (the number of dots in a certain interval), the human eye has a uniform density on the background of the document. It appears to be shaded and the presence of information is inconspicuous.
(1-3) Keeping the dot pattern to which symbols are assigned and the dot pattern to which symbols are not assigned in secret makes it difficult to decipher the embedded information.
(1-4) A pattern representing information is a collection of fine dots that are embedded as a background of a document, so that even if an embedding algorithm is disclosed, it is difficult to falsify embedded information in a printed document. .
(1-5) Since the embedded signal is detected based on the difference in the direction of the wave (change in shading) (because detailed detection is not performed in units of one pixel), the printed document is stable even if there is some dirt, etc. Information detection can be performed.
(1-6) Since the same information is repeatedly embedded and information is restored using all of the repeatedly embedded information at the time of detection, the signal portion is hidden by large font characters or the paper is dirty. Even if partial information loss occurs, the embedded information can be taken out stably.
(1-7) Further, signal units that cannot be detected accurately due to character pixels or dirt existing in the document image are excluded from the target of the majority operation in advance, so that error correction, information extraction, etc. by the majority operation are performed. Accuracy can be improved. Note that the time required for the process for obtaining the unit determination accuracy is extremely short compared with the time required for the data code reconstruction process. It is not much different from the case of doing.

(Second Embodiment)
In the data code reconstruction process performed by the watermark information detection apparatus according to the second embodiment, an error detection / correction code (for example, a Hamming code or a BCH code) is used when encoding data obtained from a watermarked document image. Is different from the data code reconstruction process (S404) of the first embodiment in that an encoding format that can determine whether the information is correct or incorrect is used. Note that it is possible to use a method of adding a checksum or CRC16 (CRC16-CCITT or the like) to the data string. Hereinafter, differences from the first embodiment will be described in detail, but the other points are substantially the same, and detailed description thereof will be omitted.

In the second embodiment, when the threshold T of the unit determination accuracy Pu is obtained, a plurality of thresholds T are calculated by using a plurality of values α existing within a predetermined range, and each threshold T is used. Perform bit confidence calculation. Each threshold value T (α _i ) is expressed by the following formula. For example, the value α is “{0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0}. It has a value such as “”.

  Hereinafter, specific processing such as data reconstruction processing according to the second embodiment, which is different from the first embodiment, will be described with reference to FIG. Is substantially the same as the bit certainty calculation of the first embodiment, and a detailed description thereof will be omitted.

As shown in FIG. 23, first, a threshold value T (α _i ) is calculated (S901). It is assumed that the initial value of i is, for example, i = 1. Next, the bit certainty calculation described in the first embodiment is performed using the threshold value T (α _i ) obtained from the above formula (S903). An error detection code, an error detection checksum, CRC16, or the like assigned on the watermark embedding side (watermark information embedding device) to the data code corresponding to the threshold T (α _i ) obtained from the bit certainty calculation Is used to perform error detection (S905).

  If an error is detected by executing the error detection (S907), the data code is discarded (excluded) as an error data code. That is, by discarding an incorrect data code, only the correct data code exists and only the correct data code can be obtained.

More specifically, the subsequent processing after the error detection will be described. In this case, _{i of} T (α _i ) is incremented by 1 (i = 1, 2, 3,...), A threshold T (α _{i + 1} ) is obtained, and unit determination is performed. Error detection is performed as a result of the bit certainty calculation for the signal detection result satisfying the relationship of accuracy Pu ≧ T (α _i ) (S905). If more errors are detected by the error detection process (S907), the error data code is eliminated as described above. Thereafter, i is further incremented by 1 to obtain a threshold value (α _{i + 2} ) (S901). Thereafter, as long as an error is detected, the above series of processing continues until i is incremented and the value α is out of range. Note that when a data code for which no error is detected is obtained for the first time, the data code may be output, and the remaining processing such as bit certainty calculation may be omitted.

Also, based on the obtained threshold value T (α _i ), the bit certainty factor is calculated, and after the error detection process, the data code is such that no error is detected (S907), and the threshold value T (α _i ) is recorded in the document A threshold value suitable for the image is dynamically determined (S909). That is, the unit determination accuracy Pu is obtained by executing the error detection process each time while sequentially changing the threshold T (α _i ) so that the maximum likelihood determination, the maximum likelihood method, and the like can be exemplified, and gradually the appropriate unit determination accuracy Pu. Is stipulated. This is the end of the description of the series of data code reconstruction processes according to the second embodiment.

  As described above in detail, according to the present embodiment, the following excellent effects are obtained.

By dynamically changing the threshold value T (α _i ) and determining an appropriate threshold value, it is possible to maximize the possibility of detecting an erroneous watermark signal in any environment regardless of the situation on the printed document or document image. The limit can be reduced.

Conventionally, the ratio between a signal unit indicating a correct bit value and a signal unit indicating an incorrect bit value is not constant on a document image due to fluctuations in printing or scanning, paper quality, dirt, and the like. In other words, it is very dependent on the dirt and paper quality. Therefore, if the threshold value T (α _i ) is set to a fixed value, a signal unit indicating a correct bit value may be discarded as an erroneous signal unit as noise, or conversely, an incorrect signal unit may be correct. In some cases, it was determined as a signal unit.

  Next, in the first embodiment and the second embodiment, the series of processes described above can be performed by dedicated hardware or software. When a series of processing is performed by software, a program constituting the software is installed in a general-purpose computer or a microcomputer.

  The program can be recorded in advance on a hard disk or ROM as a recording medium built in the computer.

  Alternatively, the program is not limited to a hard disk drive, but a removable recording medium such as a flexible disk, a CD-ROM (Compact Disc Read Only Memory), an MO (Magneto Optical) disk, a DVD (Digital Versatile Disc), a magnetic disk, and a semiconductor memory. Can be stored (recorded) temporarily or permanently. Such a removable recording medium can be provided as so-called package software.

  Here, in this specification, the processing steps for describing a program for causing a computer to perform various processes do not necessarily have to be processed in time series in the order described in the flowchart, but in parallel or individually. This includes processing to be executed (for example, parallel processing or processing by an object).

  As mentioned above, although preferred embodiment of this invention was described referring an accompanying drawing, this invention is not limited to this example. It is obvious for those skilled in the art that various changes or modifications can be envisaged within the scope of the technical idea described in the claims, and these are naturally within the technical scope of the present invention. It is understood that it belongs.

  The watermark information detection according to the present embodiment has been described by taking an example in which a printed document in which watermark information is embedded is scanned, a document image is obtained, and the watermark information is detected from the document image. For example, when all processing is performed on an electronic medium instead of a printed document, even when the watermark information is detected from a document image in which the watermark information is embedded without scanning the printed document. Good.

  The present invention is applicable to a technique for detecting confidential information from a printed document including confidential information or a document image including confidential information.

It is explanatory drawing which shows an example of the outline of the process result of filtering. It is explanatory drawing which shows an example of the outline of the data reconstruction process concerning the former. It is explanatory drawing which shows the structure of the watermark information embedding apparatus and watermark information detection apparatus concerning 1st Embodiment. 4 is a flowchart showing a processing flow of a watermark image forming unit 12. It is explanatory drawing which shows an example of a watermark signal, (1) has shown the unit A, (2) has shown the unit B. It is sectional drawing which looked at the change of the pixel value of Fig.5 (1) from the direction of arctan (1/3). It is explanatory drawing which shows an example of a watermark signal, (3) shows the unit C, (4) shows the unit D, (5) shows the unit E. It is explanatory drawing of a background image, (1) shows the case where unit E is defined as a background unit and this is used as the background of a watermark image arranged without gaps, and (2) is in the background image of (1) An example in which the unit A is embedded is shown, and (3) shows an example in which the unit B is embedded in the background image of (1). It is explanatory drawing which shows an example of the symbol embedding method to a watermark image. 5 is a flowchart illustrating a method for embedding confidential information in a watermark image. 4 is a flowchart showing a processing flow of a watermark detection unit 32. It is explanatory drawing which shows an example of a watermarked document image. It is explanatory drawing which expanded and showed a part of FIG. 4 is a flowchart showing a processing flow of a watermark detection unit 32. FIG. 15 shows an example of the input image (FIG. 15 (1)) and the input image (FIG. 15 (2)) after setting the unit pattern separation positions. It is explanatory drawing which showed an example of the area | region corresponding to the unit A in an input image. It is sectional drawing which looked at FIG. 16 from the direction parallel to the propagation direction of a wave. It is explanatory drawing explaining the method to determine whether the symbol unit embedded in the unit pattern U (x, y) is the unit A or the unit B. FIG. It is explanatory drawing which shows an example of information restoration It is explanatory drawing which shows an example of the decompression | restoration method of a data code. It is explanatory drawing which shows an example of the decompression | restoration method of a data code. It is explanatory drawing which shows an example of the decompression | restoration method of a data code. It is explanatory drawing which shows an example of the decompression | restoration method of a data code. It is explanatory drawing which shows an example of the decompression | restoration method of a data code. It is explanatory drawing which shows an example of the decompression | restoration method of a data code. It is a flowchart which shows an example of the outline of the data reconstruction process of 2nd Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Watermark information embedding apparatus 11 Document image formation part 12 Watermark image formation part 13 Watermarked document image composition part 14 Output device

Claims (18)

A watermark information detection device that extracts confidential information from an image:
An image input unit for inputting the image and obtaining an input image;
Extracting the embedded pattern by filtering the input image using a detection filter prepared for the same type as the pattern that reacts to one or more types of patterns embedded in the input image; A pattern determination unit for determining and determining a symbol corresponding to the pattern;
An accuracy calculation unit that calculates the accuracy of the symbol determination for each pattern embedded in the input image based on an output value for each pattern obtained by the detection filter ;
Comparing the calculated determination accuracy with a predetermined threshold value, selecting a pattern corresponding to the determination accuracy indicating a value larger than the threshold value, and decrypting the confidential information using the selected pattern ; Prepared,
The information decoding unit further executes a certainty factor operation for determining a symbol value constituting the confidential information encoded from one or more symbol values corresponding to the selected pattern , and based on the result, the confidentiality calculation is performed. A watermark information detection apparatus for decoding information. Prepare one or more types of identifiable patterns, give at least one symbol to one pattern, and place one or more types of the patterns generated by encoding confidential information An image input unit that reads the image in which the confidential information is embedded as an input image;
To detect the pattern from the input image, it performs filtering of the input image for detection filter responsive to the pattern prepared by the same kind as the pattern, in the image region of the input image, by the detection filter A pattern determining unit that determines and determines a symbol of the pattern based on an output value obtained;
A probability calculating unit for calculating for each pattern embedded the accuracy of symbol decision in the input image based on the output value of each pattern obtained by the detection filter;
The determination accuracy calculated by the previous Ki確 calculator with a predetermined threshold value, select the pattern corresponding to the determination accuracy indicating a value greater than the threshold value, decoding the secret information using the selected pattern An information decoding unit to perform;
With
The information decoding unit further executes a certainty factor operation for determining a symbol value constituting the confidential information encoded from one or more symbol values corresponding to the selected pattern , and based on the result, the confidentiality calculation is performed. A watermark information detection apparatus for decoding information. 3. The watermark information detection apparatus according to claim 1, wherein in the certainty factor calculation, a majority vote calculation for taking a majority vote is performed on the selected pattern. 4. The magnitude of the cumulative value of a plurality of filter output values obtained by accumulating the filter output values for each filter is compared with the selected pattern in the certainty factor calculation. 2. The watermark information detection apparatus according to 2. 3. The certainty factor calculation compares the magnitudes of accumulated values of a plurality of determination accuracy obtained by accumulating determination accuracy for each filter with respect to the selected pattern. The watermark information detection apparatus described. The watermark information detection apparatus according to claim 1, wherein the determination accuracy is a maximum value among output values obtained by filtering using one or more types of the detection filters. When two or more types of detection filters are used , the determination accuracy is a difference between output values obtained by filtering with at least two of the two or more types of detection filters. The watermark information detection apparatus according to claim 1, wherein the watermark information detection apparatus is calculated using a value. A unit pattern having at least one or more patterns formed by encoding the confidential information as a unit is repeatedly embedded in the input image a plurality of times. , 4, 5, 6, or 7. The watermark information detection device according to any one of items 7 to 7. The information decoding unit calculates a threshold value of the determination accuracy using the calculated determination accuracy of one or more, 2, 2, 3, 4, 5, 6, 7, Or the watermark information detection apparatus of any one of 8 items | items. The watermark information detection apparatus according to claim 9, wherein the threshold value of the determination accuracy is calculated and held in advance before selection of the pattern by the information decoding unit. 11. The watermark information detection apparatus according to claim 9, wherein the threshold value of the determination accuracy is an average value of the calculated one or more determination accuracy. The watermark information detection apparatus further includes an error detection unit,
The error detection unit detects an error with an error detection code for the confidential information decrypted by the information decryption unit. The watermark information detection apparatus according to any one of 8, 9, 10, or 11. The information decoding unit has a first threshold and a second threshold different from each other as the predetermined threshold used for comparison with the determination accuracy,
The error detection unit performs error detection on the pattern selected by the information decoding unit as a result of the comparison between the determination accuracy and the first threshold, and as a result, detects an error,
The information decryption unit compares the determination accuracy with the second threshold value and performs the selection of the pattern again, and further executes the certainty factor calculation for the selected pattern to obtain the confidential information . The watermark information detection apparatus according to claim 12 , wherein the watermark information detection apparatus performs decoding. The information decoding unit is configured to error by the error detecting unit is no longer detected, characterized in that said repeatedly executing the certainty calculation after the error detection by the error detection unit, either one of claims 12 or 13 wherein 2. The watermark information detection apparatus according to item 1. The information decoding unit uses the plurality of determination accuracy thresholds one by one in order one by one. 15. The watermark information detection apparatus according to any one of items 12, 13, or 14. The information decoding unit dynamically calculates the threshold value of the determination accuracy to be different from each other, characterized in that: The watermark information detection device according to any one of 11, 12, 13, 14, or 15 items. A watermark information detection method for extracting confidential information from an image comprising:
An image input step of inputting the image by an image input means to obtain an input image;
A pattern determining step of extracting one or more patterns embedded in the input image by filtering the input image by a pattern determining means , determining and determining a symbol corresponding to the pattern;
An accuracy calculation step of calculating the accuracy of the symbol determination for each pattern based on the output value of each pattern obtained by the detection filter by the accuracy calculation means;
Information by the information decoding means, a determination accuracy of the calculated and compared to a predetermined threshold, select the pattern corresponding to the determination accuracy indicating a value greater than the threshold value, to decode the secret information using the selected pattern Decryption process,
In the information decoding step, a certainty factor calculation for determining a symbol value constituting the confidential information encoded from one or more symbol values corresponding to the selected pattern is further executed, and based on the result, the confidentiality calculation is performed. A method for detecting watermark information, comprising decoding information. By preparing one or more identifiable patterns, providing at least one symbol for one pattern, and placing one or more of the patterns generated by encoding confidential information An image input step of reading an image in which the confidential information is embedded as an input image by an image input means ;
In order to detect the pattern from the input image, the pattern determination means prepares only the same type of detection filter that reacts to the pattern and performs filtering of the input image, and in the image area of the input image, wherein determining the symbol of the pattern based on the output value obtained by the detection filter, and determining the pattern determination step;
An accuracy calculation step of calculating the accuracy of symbol determination for each pattern embedded in the input image based on the output value for each pattern obtained by the detection filter by the accuracy calculation means ;
The information decoding means compares the determination accuracy calculated in the determination accuracy calculation step with a predetermined threshold, selects a pattern corresponding to the determination accuracy indicating a value larger than the threshold, and uses the selected pattern to An information decoding step for decoding information,
In the information decoding step, a certainty factor calculation for determining a symbol value constituting the confidential information encoded from one or more symbol values corresponding to the selected pattern is further executed, and based on the result, the confidentiality calculation is performed. A method for detecting watermark information, comprising decoding information.