JP2004282356A - Image processing apparatus, method, and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To make it possible to generate embedded images of high quality even when the volume of secret information is enormous in data volume by changing the format of secret information to be embedded in accordance with the embeding tolerance of an original image in which images are embedded or the sorts of images and then embeding the secret information. <P>SOLUTION: A secret information volume calculation part 101 calculates the data volume Ja of secret information A and a embedding tolerance calculation part 102 calculates the embedding tolerance Kb of an original image B. A embedding processing part 103 determines a embedding method of the secret information A on the basis of the data volume Ja and the embedding tolerance Kb. When Ja>Kb, the format (e.g. a URL (user requirement language), an address, a character code, etc.) of the secret information A is changed. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an image processing apparatus, method, and storage medium for embedding watermark information in a predetermined image or a predetermined area for the purpose of copyright protection, falsification detection, and the like.
[0002]
[Prior art]
In recent years, with the rapid development of computers and their networks, various types of information such as character data and image data have been used as digital data. These digital data can be easily copied and falsified, and copyright protection has become a major issue. One of the technologies for protecting copyright is "digital watermarking". Digital watermarking is a technique for embedding the name of a copyright holder or the ID of a purchaser in digital image data, character data, or the like in a manner that cannot be perceived by humans, and tracking unauthorized use due to illegal copying.
[0003]
In addition to copyright protection, the digital watermark is embedded with predetermined information in digital data in advance, and falsification is performed by determining whether or not the information to be extracted from the digital data has been obtained accurately. Detection can also be performed.
[0004]
When embedding digital watermark information in an original, there is a trade-off between “the difference in quality between the original and that after the digital watermark information is embedded” and “the amount of digital watermark information to be embedded”. In a natural image composed of multiple values per pixel or a halftone image composed of small values (for example, binary or ternary), a large amount of information is maintained while maintaining the quality of the original image relatively. Watermark can be embedded. However, in the case of an image composed of character information, there are many images that are visually close to binary, such as “white background and black characters”. In this case, it is difficult to embed a large amount of watermark information while maintaining the quality of the original image. As described above, when it is assumed that embedding is performed so as not to deteriorate the quality of the original image, the amount of information that can be embedded depends on the type and content of the original image that is the embedding image. However, on the other hand, there is also a request to embed predetermined confidential information regardless of the image type and content of the original image.
[0005]
For example, there has been proposed a method of embedding an Internet URL or an address capable of accessing information related to master image data (embedded image) by using a digital watermark technique (for example, see Patent Document 1). ). This method smoothly manages documents and asserts copyright, and does not consider the quality or durability of an embedded image.
[0006]
There is also a method of calculating the amount of information that can be embedded in a predetermined block unit of an embedded image based on a digital watermark embedding method based on a density pattern method, and increasing the embedding amount of the entire image as compared with the related art (for example, See Patent Document 2).
[0007]
[Patent Document 1]
JP 2001-75867 A
[Patent Document 2]
JP 2001-309150 A
[0008]
[Problems to be solved by the invention]
However, the technique of Patent Document 2 is limited to embedding by the density pattern method, and can strictly calculate the embedding possible amount of the entire image. However, when the amount of confidential information to be embedded exceeds the data amount, Cannot respond.
[0009]
The present invention has been made in view of the above problems,
An object of the present invention is to change the form of the concealment information to be embedded according to the allowable amount of embedding of the original image to be embedded or the type of the image, and to embed the confidential information. When the embedding information is image data such as a bitmap, the image itself is compressed, and the compression ratio is changed depending on the image type, so that the embedding information can be changed depending on the image type of the embedded image. Another object of the present invention is to provide an image processing apparatus, method, and storage medium that enable a receiver to grasp the entire contents of embedded information.
[0010]
[Means for Solving the Problems]
The present invention has an embedding means for embedding information A to be concealed in an arbitrary image B so as not to be recognizable, and means for determining the amount of information that can be embedded in the image B. Is characterized in that the form of the embedded data is changed according to
[0011]
According to the inventions described in claims 1, 24, and 25, by determining in advance the allowable amount of embedding of an image to be embedded, that is, the amount of embeddable information of the image itself, it becomes possible to embed all data of confidential information. It is possible to avoid. Further, by embedding information within the allowable amount of embedding of the embedded image, it is possible to provide a high-quality embedded image.
[0012]
According to the second aspect of the present invention, the image content of the embeddable image is determined, and the process of determining the area is automatically performed. It is possible to make a determination.
[0013]
According to the third aspect of the present invention, by issuing a warning, it is possible to inform the user in advance that all data cannot be embedded, and it is possible to avoid unnecessary processing in advance.
[0014]
According to the fourth aspect of the present invention, the embedding target can be reduced by embedding not the actual data of the confidential information but the storage destination address of the confidential information. That is, a high quality watermark image can be provided, and the embedding process can be performed at high speed.
[0015]
According to the fifth aspect of the present invention, the embedding target can be reduced by embedding not the actual data of the secret information but the data obtained by compressing the secret information. In other words, a high quality watermark image can be provided, and the embedding process can be performed at high speed.
[0016]
In the invention according to claim 6, for example, when the confidential information is bitmap data having a large data amount, it may not be possible to embed all data. In this case, if the embedding process is performed as it is, only the middle of the data can be embedded, and the whole image cannot be grasped at the time of extraction. Therefore, the confidential information is reversibly and irreversibly compressed at a compression ratio according to the allowable amount of embedding. This makes it possible to grasp the entire image of the confidential information.
[0017]
According to the seventh aspect of the invention, the embedding target can be reduced by embedding a character code instead of actual data of confidential information. In other words, a high quality watermark image can be provided, and the embedding process can be performed at high speed.
[0018]
According to the eighth aspect of the invention, by embedding bitmap data instead of actual data of confidential information as an embedding target, it becomes possible to provide an embedded image with high durability.
[0019]
According to the ninth aspect, an allowable amount is set in advance for each image type (character, photo, print, etc.). By using the set allowable amount of the image type designated by the user, the process of determining the allowable amount can be omitted. In addition, the entire process can be performed at a high speed.
[0020]
According to the tenth aspect of the present invention, the allowable amount is set in advance for each image type (character, photo, print, etc.), and the image type of the input image B is automatically determined. It is possible to avoid erroneous determination that can be made when giving an instruction. Thereby, it is possible to generate an appropriately high-quality embedded image. According to the present invention, high-quality embedded images can be generated by embedding confidential information in various forms in an appropriate form.
[0021]
According to the twenty-third aspect, when the image type is determined to be a mixed character / photo image, the amount that can be embedded varies depending on the ratio of those areas that are actually arranged. Thus, by detecting the ratio at the time of determination and using the ratio to determine the amount of embedding, highly accurate embedding processing can be performed.
[0022]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be specifically described with reference to the drawings.
[0023]
(Example 1)
First Embodiment As a first embodiment, an embodiment using a method of automatically determining an embeddable information amount (hereinafter, an allowable amount Kb) of an original image B (hereinafter, an original image B) which is an embedded image will be described.
[0024]
FIG. 1 shows the configuration of Embodiment 1 of the present invention. The secret information amount calculation unit 101 calculates the data of the secret information A and the data amount Ja thereof. The allowable embedment calculating unit 102 calculates the data of the original image B and the allowable embedment Kb of the original image B. Each output is input to the embedding processing unit 103, and here, the embedding method of the confidential information A is determined based on the calculated values of Ja and Kb.
[0025]
First, the embedding allowable amount calculation unit 102 will be described. Here, the approximate allowable amount of embedding of the original image B is determined. For example, the redundancy of the image differs depending on the image content such as characters, photographic paper, and photographs. In the case of a general character image, a portion other than the character is expected to be white or have a constant density. At this time, the redundancy is extremely low. That is, since the area in which information can be embedded is concentrated on the edge portion of the character, the embedding amount of the entire image is reduced. This does not change much even if the image becomes multi-valued. On the other hand, photographic paper and photographic images generally have a high degree of redundancy, and when the value is multi-valued rather than binary, the redundancy is further increased. Therefore, it can be expected that the embedding allowance will also increase.
[0026]
A number of known techniques have been reported as a method for judging image contents. However, "DCT coding method using adaptive quantization-operation to digital color copier-", Kazuyuki Murata (Journal of the Institute of Image Electronics Engineers of Japan) The method described in Vol. 20, No. 5 (1991)) utilizes a DCT (Discrete Cosine Transform) coefficient, which is one of the frequency transforms, to determine the character, halftone dot, Are divided into three regions.
[0027]
Utilizing such a separation technique, the separation result is divided into a character area, a photograph / halftone dot area, and an impossible area, and the embedding amount per area is set for each area. FIG. 2 shows an example of the region separation result. When the original image B is the image 200, the area separation is performed as in the areas 201, 202, and 203 using the above-described separation method. An area 201 is an enlarged view of a result of determination of a character image, and one square (block) indicates one area. When the determination is made with 8 × 8 pixels per area, a set of 8 × 8 pixels is obtained. An area 204 indicates the result of separation of the area 201. The edge portion of the character is determined to be a character region, and the solid portion and the periphery of the character are determined to be non-embeddable regions. Although the area 202 is a rectangle of a photographic image, almost the entire surface is determined to be a photographic / dot area. Since the region 203 is a white background portion, it is determined to be an unacceptable region. The “impossible region” is determined as a non-embeddable portion where the density of a photograph or photographic paper is small, or a portion with low redundancy such as a white background portion or a solid portion.
[0028]
Next, the setting of the embedding amount of each area will be described. For example, when the bit replacement type embedding is performed as an embedding method, 1/0 code is embedded by replacing data near a character edge in a predetermined bit plane of an 8 × 8 block determined as a “character area”. At this time, the number of bits that can be embedded per area block depends on the amount of the edge portion of the character, and therefore varies depending on the block. Therefore, it is better to estimate the number of bits to be embedded in the character area to be set in advance with a small number in consideration of this point. On the other hand, in the "halftone dot area", all the predetermined bit planes are similarly replaced with the confidential information A. At this time, if the separated area size is 8 × 8 pixels, 64 bits can always be embedded. Further, since the “impossible area” has a small allowable amount of embedding, no embedding is performed.
[0029]
For example, when Lbit is set for a character area, Mbit is set for a photograph / halftone area, and Nbit is set for an unusable area, the relationship is as follows.
M>L> N (bit).
[0030]
The actual number of bits is set from an empirical value based on experiments. Since the character area is embedded in the edge portion of the character, the embedding amount is relatively small, and the halftone / photo area is set to be relatively large. Since the unusable area is an area where image deterioration is conspicuous, embedding is not performed.
[0031]
FIG. 11 shows an example of a method for calculating the allowable embedding amount Kb. For example, it is assumed that the area ratio of a certain image is 50% for a character area, 30% for a photograph / halftone area, and 20% for an unusable area. In this case, the set embedding amount of each area is set to a character area L = 10 bits, a photograph / halftone dot area M = 64 bits, and an impossible area N = 0 bits.
[0032]
Pbit is obtained as the allowable number of bits for embedding of the entire character area by “total number of areas (width / 8 * height / 8) * character area ratio (50/100) * embedding amount per area (10 bits)”. Similarly, the photograph / halftone dot area is Q bits, and the embedding allowance of the entire image is (P + Q) bits.
[0033]
The above-described separation technique and embedding technique are examples, and the present invention is not limited to the above-described technique, and other techniques may be used. For example, as a method of embedding in a multi-valued still image, a method using patchwork, a method using frequency conversion to embed in a frequency domain, a pixel replacement type, a bit replacement type, and a run length of black pixels for a binary image are used. A known technique such as replacement may be used.
[0034]
FIG. 3 is an embedding processing flow of the confidential information A in the embedding processing unit 103. The embedding processing unit 103 stores the data (hereinafter, Da1) of the confidential information A of the first form, the information amount Ja1 (hereinafter, Ja1), the original image B, and the embedding allowable amount Kb (hereinafter, Kb) of the original image B. ) Is entered. The embedding processing unit 103 determines in what form the secret information A is embedded based on the relationship between Ja1 and Kb.
[0035]
If the relationship of “Ja1 <Kb” is satisfied with the first form in which the confidential information A is initially input (Yes in step 301), the embedding process is performed with the original first form of Da1 (step 301). 302). However, when “Ja1> = Kb” (No in step 301), the data form (Da1) of the confidential information A is changed to data of the second form (Da2) (step 303), The amount Ja2 is calculated (step 304), and if “Ja2 <Kb” is satisfied (Yes in step 305), embedding is performed in the second mode (Da2) (step 302). However, if “Ja2> = Kb” (No in step 305), a warning is issued to notify the user that embedding is not possible (step 306), and the process ends.
[0036]
(Example 2)
As a second embodiment, a case where the confidential information A is character information will be described as an example. In the process of “changing the form of the confidential information A from the first form to the second form” in step 303 in FIG. 3, in the first embodiment, an example in which the embedding is performed based on only the confidential information amount (Ja1) is performed. As described above, here, an embedding method in consideration of the form of secret information will be described.
[0037]
When the confidential information A is character information, there are two possible forms of the confidential information A as a transmission means, such as a character code or a bit-mapped image of a character. In the general idea of watermarking, when resistance to attack is considered, it can be said that embedding an image is more resistant. In a character code, if even one bit is changed, the code itself is changed, so that it is extracted as a completely different character. However, if the image is embedded as an image, even if it is changed by 1 bit, the character as the image can be decoded. Therefore, when there is no problem in image quality, embedding in the form of a bitmap can provide an embedded image with higher durability.
[0038]
FIG. 4 is a flowchart of processing for embedding confidential information according to the second embodiment. Here, an example is given in which the bitmap form is prioritized over the character code form for the above-described tolerance.
[0039]
If the first form (Da1) of the confidential information A at the time of input is bitmap data (Yes in step 401), and if "the information amount: Ja1 <Kb" holds (Yes in step 402), the bitmap Embedding is performed in the form of data (step 403). However, if it is a character code (No in step 401), the character code is expanded into a temporary storage unit such as a memory in a bitmap manner, and the confidential information A is changed to bitmap data in the second form (Da2) ( Step 404).
[0040]
The information amount Ja2 of the second mode (Da2) is calculated (step 405), and “(data amount of bitmap of secret information A): Ja1> = Kb” or “(character code data of secret information A) (Amount): Ja2> = Kb ”(No in Steps 402 and 406), only the embeddable amount is embedded, or a warning is issued to the user to notify that embedding is impossible (Step 407). ). Alternatively, as shown in the flowchart of FIG. 5, when “Ja1> = Kb” (Yes in step 501), data compression processing is performed on the bitmap data of Da1 (step 502). If “Ja1 <Kb” is not satisfied by one compression, recompression is performed by changing the compression rate until the data amount becomes embeddable.
[0041]
(Example 3)
In the second embodiment, the embedding in consideration of the tolerance has been described. In the third embodiment, a method in which the image quality after the embedding is considered will be described. As for the quality of the image, the lower the amount of information to be embedded, the more the deterioration of the original image B is suppressed. Therefore, in the third embodiment, priority is given to reducing the data amount, and priority is given to the character code form over the bitmap form.
[0042]
FIG. 6 is a flowchart of processing for embedding confidential information according to the third embodiment. Similarly to the second embodiment, when the confidential information A is character information and the first form (Da1) of the confidential information A at the time of input is bitmap data (Yes in step 601), the form of Da1 is changed to a character. The data is converted into Da2 (step 602). The conversion here requires some kind of processing such as OCR for converting bitmap data into character codes. Next, the information amount Ja2 of (Da2) is calculated (step 603), and when “Ja2 <Kb” is obtained (Yes in step 604), embedding is performed (step 605). If the answer is No in Step 604, a warning is issued (Step 607), and the process ends. If No in step 601, proceed to step 606. If "Ja1 <Kb" (Yes in step 606), embed (step 605). If no in step 606, issue a warning (step 607) and end. I do.
[0043]
FIG. 7 is a processing flowchart for avoiding embedding failure. The processing up to step 704 is the same as in FIG. If the determination in step 704 is No, a warning is not immediately issued, and in step 706, if there is a URL (Uniform Resource Locator) for accessing the confidential information A, or an address specifying the confidential information A itself, if any, The form can be changed to that shown (Da3) and embedded (step 705).
[0044]
In the above example, an example in which the confidential information A is character information has been described. However, the same processing can be performed for image information. In this case, if the bitmap data is embedded as it is, the embedding ends when the allowable amount of the image B is exceeded. As a result, since only a part of the confidential information A is embedded, the receiver may not be able to understand the entire image of the confidential information A.
[0045]
FIG. 8 is a processing flowchart when the confidential information is image information. Basically, the form of the confidential information A is a form such as bitmap data. If the information amount becomes Ja1> = Kb (No in step 801), the process proceeds to step 803 to access the confidential information A. If there is an URL for specifying the URL or the secret information A itself, the form is changed to indicate the address (Da2) and embedded in step 802. Alternatively, as shown in the processing flowchart of FIG. 5, embedding may be performed after the secret information A itself is reversibly or irreversibly compressed to reduce the amount of embeddable information. This allows the recipient to know the entirety of the confidential information A, and also maintains the quality of the image B.
[0046]
As in the above-described first to third embodiments, the most accurate method is to perform area separation in units of small blocks of an image and calculate the total embedding allowable amount, but the cost and processing time of this processing itself also increase. Will do. Thus, as described below, the user can specify the type of image, or the image type determination unit can automatically determine the type of image and determine the embedding amount based on the result. As described above, by simply performing the image type determination, the area separation processing can be simplified, and the processing for calculating the total allowable embedding amount itself can be omitted.
[0047]
(Example 4)
Fourth Embodiment As a fourth embodiment, a method for determining the embeddable information amount (Kb) of the original image B, which is an embedded image, after determining the image type of the image B will be described.
[0048]
FIG. 9 illustrates a first configuration according to the fourth embodiment. FIG. 8 shows a configuration example in which the user designates the type of image, and a document type input unit 904 is added to the configuration of FIG. For example, the user instructs the type of the image by his / her own judgment from the image type input unit 904, which is a selection means such as “character / photo / gravure”. When the image is designated as a character image, it is determined by assuming the number of bits of the entire image by using the number of bits per character area (Lbit) previously set. However, since there are actually many blank portions in the character image, the ratio of the blank region (Nbit) is provisionally set, and a value is set in consideration of the ratio.
[0049]
In this process, it is more efficient to set in advance a value as an image such as a character image = Rbit, a photographic image = Jbit, and a gravure (printing paper) image = Kbit, because the process is simplified and the efficiency is higher.
[0050]
FIG. 10 shows a second configuration according to the fourth embodiment. FIG. 10 shows an example in which an image type determination unit 1004 is provided to automatically determine the type of an image. Various image type determination techniques have been proposed as known techniques. The determination may be made by searching all data of the input image B to determine the image type, or by searching only a predetermined part of the image. The amount of information (Kb) that can be embedded is determined based on the image type determined by these techniques in the same manner as in the example in which the user determines the image type described with reference to FIG.
[0051]
In the present invention, the form of the confidential information A is determined so as to be optimal for the determined various image types in consideration of image quality and embedding resistance. Further, when it is determined that the image type is a mixed image, the amount that can be embedded actually changes depending on the mixing ratio. Therefore, it is appropriate to detect the mixture ratio and determine the possible embedding amount according to the mixture ratio. As a method of detecting the mixture ratio, it can be calculated by the area separation method as described in the first embodiment.
[0052]
In the above-described embodiment, the embedded image may be image data obtained by reading a paper original from a scanner, or may be created as electronic data. Also, confidential information may be input as character information from the input means on the embedding device, may be information stored in a predetermined place in advance, may be image data read from a scanner, or may be electronic data. It may be created.
[0053]
【The invention's effect】
As described above, according to the inventions described in claims 1, 24, and 25, the embedding allowable amount of an embedded image, that is, the amount of information that can be embedded in the image itself is determined in advance, so that the total amount of confidential information can be determined. It becomes possible to avoid the fact that data cannot be embedded. Further, by embedding information within the embedding allowable amount of the embedded image, it becomes possible to provide a high quality embedded image.
[0054]
According to the second aspect of the present invention, the image content of the embeddable image is determined, and the process of determining the area is automatically performed. The capacity can be determined.
[0055]
According to the third aspect of the present invention, by issuing a warning, it is possible to notify the user in advance that all data cannot be embedded, and it is possible to avoid unnecessary processing in advance. Become.
[0056]
According to the fourth aspect of the present invention, the embedding target can be reduced by embedding not the actual data of the confidential information but the storage destination address of the confidential information. That is, a high quality watermark image can be provided, and the embedding process can be performed at high speed.
[0057]
According to the fifth aspect of the present invention, the embedding target can be reduced by embedding not the actual data of the confidential information but the compressed data of the confidential information. In other words, a high quality watermark image can be provided, and the embedding process can be performed at high speed.
[0058]
According to the invention of claim 6, for example, when the confidential information is bitmap data having a large data amount, it may not be possible to embed all data. In this case, if the embedding process is performed as it is, only the middle of the data can be embedded, and the whole image cannot be grasped at the time of extraction. Therefore, the confidential information is reversibly and irreversibly compressed at a compression ratio according to the allowable amount of embedding. This makes it possible to grasp the entire image of the confidential information.
[0059]
According to the seventh aspect of the present invention, the embedding target can be reduced by embedding not the actual data of the confidential information but the character code. In other words, a high quality watermark image can be provided, and the embedding process can be performed at high speed.
[0060]
According to the eighth aspect of the present invention, by embedding bitmap data instead of actual data of confidential information as an embedding target, it becomes possible to provide an embedded image with high durability.
[0061]
According to the ninth aspect, an allowable amount is set in advance for each image type (character, photo, print, etc.). By using the set allowable amount of the image type designated by the user, the process of determining the allowable amount can be omitted. In addition, the entire process can be performed at a high speed.
[0062]
According to the tenth aspect, an allowable amount is set in advance for each image type (character, photo, print, and the like), and the image type of the input image B is automatically determined. It is possible to avoid erroneous determination that can be made when the user gives an instruction. Thereby, it is possible to generate an appropriately high-quality embedded image.
[0063]
According to the present invention, high-quality embedded images can be generated by embedding confidential information in various forms in an appropriate form.
[0064]
According to the twenty-third aspect, when the image type is determined to be a text / photo mixed image, the amount that can be embedded varies depending on the ratio of those areas that are actually arranged. Thus, by detecting the ratio at the time of determination and using the ratio to determine the amount of embedding, highly accurate embedding processing can be performed.
[Brief description of the drawings]
FIG. 1 shows a configuration of a first exemplary embodiment of the present invention.
FIG. 2 shows an example of a region division result.
FIG. 3 is a flowchart of an embedding process according to the first embodiment.
FIG. 4 is a flowchart of an embedding process according to the second embodiment.
FIG. 5 is a processing flowchart for avoiding embedding failure according to the second embodiment.
FIG. 6 is a flowchart of an embedding process according to a third embodiment.
FIG. 7 is a processing flowchart for avoiding embedding failure according to the third embodiment.
FIG. 8 is a processing flowchart according to the third embodiment when confidential information is image information.
FIG. 9 shows a first configuration according to a fourth embodiment.
FIG. 10 shows a second configuration according to the fourth embodiment.
FIG. 11 shows an example of an embedding amount calculation method.
[Explanation of symbols]
101 Secret information amount calculation unit
102 Embedding allowable amount calculation unit
103 Embedding processing unit

Claims (25)

An image processing device that embeds the secret information A in an arbitrary image B so that the secret information A cannot be recognized. The image processing device includes: a unit that determines an embeddable information amount of the image B; Means for changing the form of the secret information A and embedding the secret information A in the image B. The image processing apparatus according to claim 1, wherein the means for determining the embeddable information amount automatically determines the embeddable information amount of the image B. 2. The image processing apparatus according to claim 1, wherein a warning is issued when the secret information A exceeds the embeddable information amount. 2. The image processing apparatus according to claim 1, wherein the form of the secret information A is changed to a URL or an address for accessing the secret information A. 2. The image processing apparatus according to claim 1, wherein the form of the secret information A is changed to compressed information. The image processing apparatus according to claim 5, wherein a compression ratio of the compression information is changed according to an amount of information in which the image B can be embedded. 2. The image processing apparatus according to claim 1, wherein the form of the secret information A is changed to a character code. 2. The image processing apparatus according to claim 1, wherein the form of the secret information A is changed to bitmap data. An image processing apparatus that embeds the secret information A in an arbitrary image B so that the secret information A cannot be recognized, and includes a unit that specifies a type of the image B, and an embedment of the image B based on the type of the specified image B. An image processing apparatus comprising: means for determining an information amount; and means for changing a form of the secret information A and embedding the secret information A in the image B according to the determined information amount that can be embedded. An image processing device that embeds the secret information A in an arbitrary image B so that the secret information A cannot be recognized. The image processing device determines a type of the image B, and embeds the image B based on the determined type of the image B. An image processing apparatus comprising: means for determining an information amount; and means for changing a form of the secret information A and embedding the secret information A in the image B according to the determined information amount that can be embedded. When the image B is a character image and the confidential information A is image information, the form of the confidential information A is changed to a URL for accessing the confidential information A or an address or compressed information. The image processing device according to claim 9. 11. The image processing apparatus according to claim 9, wherein when the image B is a character image and the confidential information A is character information, the form of the confidential information A is changed to a character code. When the image B is a monochrome image and the confidential information A is image information, the form of the confidential information A is changed to a URL for accessing the confidential information A or an address or compressed information. The image processing device according to claim 9. 11. The image processing apparatus according to claim 9, wherein when the image B is a monochrome image and the confidential information A is character information, the form of the confidential information A is changed to a character code. When the image B is an image whose resolution has been reduced by the scaling process and the confidential information A is image information, the form of the confidential information A is changed to a URL or an address for accessing the confidential information A, or to compressed information. The image processing apparatus according to claim 9, wherein the image processing apparatus is changed to: 11. When the image B is an image whose resolution has been reduced by scaling processing and the confidential information A is character information, the form of the confidential information A is changed to a character code. The image processing apparatus according to claim 1. 11. The image processing apparatus according to claim 9, wherein when the image B is a photographic image and the confidential information A is character information, the form of the confidential information A is changed to bitmap data. 11. The image processing apparatus according to claim 9, wherein, when the image B is a photographic image and the confidential information A is image information, the confidential information A is changed to the confidential information A or compressed information. . 11. The image processing apparatus according to claim 9, wherein when the image B is a color image and the confidential information A is character information, the form of the confidential information A is changed to bitmap data. 11. The image processing apparatus according to claim 9, wherein when the image B is a color image and the confidential information A is image information, the confidential information A is changed to the compressed information or the form of the confidential information A. . 10. When the image B is an image whose resolution has been increased by a scaling process and the confidential information A is character information, the form of the confidential information A is changed to bitmap data. An image processing apparatus according to claim 10. 10. When the image B is an image whose resolution has been increased by the scaling process and the confidential information A is image information, the confidential information A is changed to the form of the confidential information A or to compressed information. Or the image processing apparatus according to 10. When the image B is an image in which a character area and a photograph area are mixed, the means for determining the embeddable information amount of the image B determines the embeddable information amount according to the mixed ratio of each area. The image processing apparatus according to claim 9, wherein: An image processing method for embedding confidential information A in an arbitrary image B so as to be unrecognizable, comprising: determining an amount of embeddable information of the image B; Changing the form of the confidential information A and embedding it in the image B. A computer-readable recording medium on which a program for causing a computer to implement the image processing method according to claim 24 is recorded.

Images