JP2011124663A - Image processing apparatus and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To appropriately verify tampering even when resolution conversion is executed to image data. <P>SOLUTION: The image processing apparatus includes: an image input part for inputting image data; a division part for dividing the input image data into a plurality of blocks; an extraction part for generating a feature amount for each of the image data divided in the division part; a generation part for generating verification data corresponding to the feature amount extracted in the extraction part; and an output part for adding information indicating the division method of the division part and the verification data to the input image data and outputting them. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an image processing apparatus and its control method for guaranteeing the originality of digital data.

  There is a concern that image data taken with a digital camera or the like may be altered using image editing software. As one method for solving this problem, a method disclosed in Patent Document 1 is known. According to this document, camera-specific secret information is incorporated in advance in a digital camera, and signature processing is performed on image data using the secret information in the digital camera. Then, after shooting, verification can be performed using the signature information. By the way, such a signature technique is generally regarded as falsification when image data is changed even a little. Therefore, resolution conversion of image data is not permitted.

US Patent No. 5,499,294

  The present invention has been made in view of such problems, and provides a technique capable of appropriately tampering verification even when resolution conversion is performed on image data.

  In order to solve the above problems, according to one of the image processing apparatuses of the present invention, an image input unit that inputs image data, a dividing unit that divides the input image data into a plurality of blocks, and the dividing unit An extraction unit that generates a feature amount for each of the image data divided by the above, a generation unit that generates verification data corresponding to the feature amount extracted by the extraction unit, and the division unit into the input image data And an output unit for adding the verification data and outputting the information.

  According to the present invention, even if resolution conversion is performed on image data, it is possible to appropriately tamper with verification.

The figure explaining the structure of an image input device The figure explaining the composition of a verification device Diagram explaining block division processing Flow chart of image input processing Flow chart of feature amount extraction processing Diagram explaining block division processing

[Example 1]
<Configuration of image input device>
FIG. 1 is a diagram illustrating an image processing apparatus (hereinafter referred to as an image input apparatus) that generates verification data of image data in the present embodiment. This image input device corresponds to, for example, a digital camera that takes an image and applies a digital signature to the image data. In particular, in the present embodiment, it is important that information indicating a method for dividing an image, which will be described later, is also provided in addition to the verification data provided to the image data when the digital signature is applied.

  The image input apparatus according to the present embodiment includes an image generation unit 11, a block division unit 12, an extraction unit 13, a verification data generation unit 14, and an image output unit 15. The block dividing unit 12 includes an input unit 16 and a dividing unit 17.

  The image generator 11 acquires image data image data I photographed by a CCD or the like. The image data I is, for example, image data representing 1600 × 1200 pixels. Hereinafter, in order to simplify the description, it will be described as a monochrome image composed of one color component. If it is applied to a color image, the processing may be repeated for the number of the color components.

  The block dividing unit 12 includes an input unit 16 and a dividing unit 17 therein. The input unit 16 receives an instruction signal indicating a division method for dividing the image data I (for example, how many blocks are divided vertically and how many are divided horizontally). This instruction is input to the input unit 16 by an operator of the image input device, for example. The dividing unit 17 divides the image data I based on the dividing method input from the input unit 16.

  FIG. 3 shows an example of a block division process applicable to this embodiment. The image 61 is an example of the image data I. Originally, about 1600 pixels × 1200 pixels is assumed, but in order to simplify the description, the description will be made with 4 × 4 pixels (256 gradations per pixel). C0 to C15 of the image 61 shown on the left in FIG. 3 are indexes indicating pixel positions. Here, the pixel value of C0 is 128, and the pixel value of C1 is 150. Similarly, C2 = 121, C3 = 90, C4 = 120,..., C14 = 115, and C15 = 155.

  The image 62 is an example of dividing the image data I into four square blocks. Here, the block dividing unit 12 divides the image 61 into four (2 × 2) blocks in accordance with a division method of “two divisions vertically and two divisions horizontally”. First, the number of block divisions (m × n) indicating how many parts the image data I is divided vertically and horizontally is input from the input unit 16. In addition, as a data structure representing the number, for example, a data string {m, n} is determined in advance. This means that there are m divisions vertically and n divisions horizontally. Next, the dividing unit 17 divides the image (the image indicated by the image data I) into m equal parts and n equal parts horizontally based on the image data I and the number of block divisions. Thereby, image data divided into m × n blocks is obtained. For example, when 61 image data I in FIG. 3 is input and the number of block divisions is m × n = 4, the blocks are divided so that 2 × 2 pixels are included in order from the upper left. As a result, four square blocks (2 × 2 pixels) shown in the image 62 of FIG. 3 are obtained. Each of the plurality of image data divided here is set as image data DI, and each image data DI is output to the extraction unit 13 in the subsequent stage.

  In the above description, the case where all the pixels of the image data I (image 61) can be equally divided has been described. In that case, the exception handling is handled as follows. For example, when the image data I (image 61) is 4 × 4 pixels and the number of block divisions is m × n = 9, division is performed as shown in an image 63. First, the number of pixels / number of divisions = 4/3 = 1.33... Is calculated in the vertical direction, and the second decimal place is rounded down to 1.3. Also in the horizontal direction, the number of pixels / number of divisions = 4/3 = 1.33 is calculated, and the second decimal place is rounded down to 1.3. Then, 3 × 3 virtual blocks are configured with one block size of 1.3 pixels in both vertical and horizontal directions. The virtual block is superimposed on the 4 × 4 pixel image 61. In this case, both ends of the virtual block do not necessarily match 4 × 4 pixels. Therefore, in this embodiment, the upper left vertex of the image 61 and the upper left vertex of the virtual block are superimposed so as to match. Next, for each of the nine blocks constituting the image 63, the blocks are interpreted as follows. First, in a portion where a certain block is arranged, if even some of the original pixels overlap, it is interpreted that the pixel is included in the block. For example, in the block arranged at the upper left of the image 63, four pixels C0 (128), C1 (150), C4 (120), and C5 (185) overlap. Therefore, the upper left block is interpreted as being composed of pixels C0 (128), C1 (150), C4 (120), and C5 (185). By interpreting the other blocks in the same manner, feature amount extraction for each block, which will be described later, can be performed appropriately.

  The extraction unit 13 generates a feature amount of the image data DI input from the block division unit 12. Here, first, all pixel values constituting a block to be processed (for example, in the case of the image 62, one block is composed of 2 × 2 pixels) are extracted. Then, for each block, an average value is calculated from all the pixel values constituting the block. The average value is the feature amount of the block. For example, in the case of obtaining the feature amount of the upper left block in the image 62 in FIG. 3, the average pixel value of this block is (128 + 150 + 120 + 185) / 4 = 145. The feature amounts of the upper right, lower left, and lower right blocks of the image 62 are 106, 139, and 163 (fractions after the decimal point are discarded), respectively. The feature amount of each block calculated here is sequentially output to the verification data generation unit 14 at the subsequent stage as a value that is the basis of the verification data of each block. For reference, when the feature amount of the image 63 in FIG. 3 is obtained, the feature amount of the upper left block is (128 + 150 + 120 + 185) / 4 = 145, similar to the feature amount of the image 62. The feature amount of the upper right block of the image 63 is an average (106) of pixel values of C2 (121), C3 (90), C6 (135), and C7 (80).

  The verification data generation unit 14 inputs the feature amount of each image data DI generated by the extraction unit 13, and generates verification data from the feature amount. Then, the generated verification data is output to the subsequent image output unit 15. As the verification data, a hash value, a MAC (Message Authentication Code), an electronic signature, or the like can be applied. In this embodiment, verification data is generated for each feature amount (image data DI) of each block. By holding the verification data for each block, there is an effect that tampering verification can be performed for each block.

  In addition, when applying MAC as verification data, the secret information for producing | generating MAC is input as signature key KS, and is utilized when producing | generating MAC. Since the signature key KS needs to be shared between the image input device and the verification device described later, common secret information is held in advance in the image input device and the verification device. On the other hand, when an electronic signature is applied as verification data, a secret key for generating an electronic signature is input as a signature key KS. The signature key KS is held in advance in the image input apparatus, and the verification data generation unit 14 uses the signature key KS as necessary. It is assumed that a public key corresponding to the signature key KS used by the verification data generation unit 14 is held in a verification device described later.

  The image output unit 15 adds two pieces of information indicating the division method applied to the division of the block division unit 12 and the verification data generated by the verification data generation unit 14 to the image data I before being divided into blocks. And output. As described above, the verification data added here is the verification data of each block, which is the same number of verification data as the number of blocks. Here, the data arrangement of the verification data is determined in advance so that which verification data corresponds to which block. In this embodiment, the information indicating the division method and the verification data are added as header information of the image data. However, the present invention is not limited to this, and it may be connected after the image data I. The image output unit 15 records the image data I including the header information in various memories or transmits it to the outside via a network.

<Processing process of image input device>
FIG. 4 is a diagram showing a series of flow from image capturing, digital signature and image data output performed by the image input apparatus described above. By the way, the operation of the image input apparatus shown in FIG. 1 is controlled based on this flow, but part of the intermediate operation can be performed by software processing.

  In particular, the processing of the block division unit 12, the extraction unit 13, and the verification data generation unit 14 uses a CPU and a memory that holds a program for performing a series of operations in FIG. 4, a memory that holds intermediate data being processed, and the like. Software processing is also possible.

  Hereinafter, the flow of the processing in FIG. 4 will be described. First, in step S71, the image generating unit 11 captures an image and acquires image data I. As described above, the image data I is image data representing, for example, 1600 × 1200 pixels. In step S72, the block dividing unit 12 divides the image data I into a plurality of blocks. In step S73, the extraction unit 13 generates a feature amount from the divided image data DI. Here, the feature amount of each block is generated. In step S74, the verification data generation unit 14 generates verification data of each block using the feature amount of each block generated by the extraction unit 13. In step S75, the image output unit 15 outputs the image data I, information indicating the division method (number of divided blocks), and verification data of each block as a set.

<Configuration of verification device>
Hereinafter, the tampering verification apparatus of the present embodiment will be described with reference to FIG. As illustrated in FIG. 2, the verification apparatus according to the present exemplary embodiment includes an image input unit 101, a block division unit 102, an extraction unit 103, and a verification unit 104. The block dividing unit 102 includes an input unit 16 and a dividing unit 17. First, the image data I output from the above-described image input device is input to the verification device as image data I ′ via the storage medium or the network. As described above, the image data I ′ is accompanied by information (number of divided blocks) indicating the above-described division method and verification data as header information. The block division unit 102 and the extraction unit 103 perform the same processes as the block division unit 12 and the extraction unit 13 described above, respectively. The verification unit 104 uses the verification data obtained from the feature amount of each block output from the extraction unit 103 to configure the input image data I ′ (more specifically, an image represented by the image data I ′). Whether each block is falsified and outputs a verification result (OK / NG). In this embodiment, since verification data is assigned to each block, when the verification result “NG” is output, information indicating which block is NG is also output. The verification process executed by the verification unit 104 must completely correspond to the verification data generation unit 14 described above. That is, when the verification data generation unit 14 generates a hash value, the verification unit 104 verifies tampering of each block using the hash value. When the verification data generation unit 14 generates a MAC, the verification unit 104 also performs verification processing using the MAC. If the verification data generation unit 14 generates an electronic signature, the verification unit 104 also performs verification processing using the electronic signature. For the verification key KS, the same secret information as the signature key KS applied by the verification data generation unit 14 is applied in the case of MAC, and the signature key KS applied by the verification data generation unit 14 is applied in the case of an electronic signature. It should be considered as the corresponding public key. Since the verification method using the MAC and the electronic signature is a technique known to those skilled in the art, a detailed description thereof will be omitted.

<Verification processing flow>
FIG. 5 is a flowchart showing the flow of verification processing executed by the verification apparatus of this embodiment. Hereinafter, the flow will be described with reference to FIG. Incidentally, the operation of the verification apparatus shown in FIG. 5 is controlled based on this flow, but part of the intermediate operation can be performed by software processing. In particular, the processing of the block division unit 102, the extraction unit 103, and the verification unit 104 may be replaced with a storage medium that holds the CPU and a program that performs the above-described operation and that holds intermediate data being processed. The present invention includes such modifications.

  First, in step S111, the image input unit 101 inputs image data I 'to be verified. In step S 112, the image data I ′ is divided into a plurality of blocks by the block dividing unit 102. Here, based on the number of divided blocks (information indicating the division method) attached to the image data I ′, the image data I ′ is divided in the vertical direction and the horizontal direction so as to be exactly the same number as the image input device. To do. Subsequently, in step S113, the extraction unit 103 extracts each feature amount of each divided block (image data DI '). In step S114, the verification unit 104 verifies whether the input image data I ′ (more specifically, each block constituting the image represented by the image data I ′) has been tampered with. In step S115, it is determined whether or not the image data I 'has been falsified (more specifically, whether or not each block constituting the image represented by the image data I' has been falsified). In step S116, if the verification is successful, the verification result “OK” is output, otherwise the verification result “NG” is output. In this embodiment, since verification data is assigned to each block, when the verification result “NG” is output, information indicating which block is NG is also output.

[Modification]
<Elements of features>
In the above embodiment, the feature amount is extracted using the average of the “pixel values”. However, the present invention is not limited to this, and a value obtained by quantizing the pixel value may be used instead of the pixel value. That is, for each block, the average of “quantized pixel values” may be used as the feature amount. Further, instead of “average” of pixel values, “dispersion” of pixel values may be used as the feature amount.

<Multiplication of verification data>
In the above embodiment, since the dividing unit 17 applies one kind of dividing method to an image of one screen (for example, the image 61), one verification data is assigned to each region in the image. State. However, for the purpose of improving the accuracy of falsification detection, it is better that two or more verification data exist in each region in the image. Therefore, as a modification, the image input apparatus divides the image data I (image 61) by a plurality of types of division methods (first division method, second division method, etc.), and corresponds to each division method. By generating the verification data, the accuracy of tampering detection by the verification device can be increased. As a specific example, the verification data A obtained by dividing the image data I (image 61) by the method of the image 62 in FIG. 6 and the verification data B obtained by dividing by the method of the image 63 are shown. Together. Then, both the verification data A and the verification data B are added as header information of the image data I and output from the image output unit 15. Therefore, the header information of the output image data I includes “division method A: {2, 2}”, “division method B: {3, 3}”, “verification data A (2 × 2 pieces of verification data). ) ”,“ Verification data B (3 × 3 verification data) ”.

<Variable block size>
In the above description, the sizes of the divided blocks are all the same, but the data amount of the verification data can be reduced if it is changed as necessary. For this purpose, for example, the size of the block (the total number of blocks) may be changed by inputting region information indicating an important region and an unimportant region from the input unit 16.

  A specific example will be described below with reference to FIG. Compared to the division method of FIG. 3, the block division method is finally determined with reference to the area information 64. That is, in the block division shown in FIG. 6, the final block is not determined only from the information indicating the block division method described in the above embodiment. In the area information 64, 1-bit information (1/0) indicating whether it is important or not important for each pixel is assigned. Here, the pixels C0, C5, C8, C9, C12, and C13 are important, and the other ten pixels are not important. In the modification shown in FIG. 6, whether the image 61 is divided into 2 × 2 or 3 × 3, post-processing based on the region information 64 exists thereafter. First, an example of dividing the image 61 into 2 × 2 will be described. It is determined whether or not at least one important pixel is included in each of the four blocks obtained by the division. Looking at the image 62, the upper left and lower left blocks are important blocks because they contain one or more important pixels. On the other hand, the upper right and lower right blocks are non-important blocks because they do not include any important pixels. In this modification, unimportant blocks are integrated as the final post-processing. Accordingly, since the upper right block and the lower right block are integrated in the four blocks of the image 62, the three blocks shown in the image 65 are the final blocks. Incidentally, since the integrated block includes 8 pixels, the above-described feature amount is an average of the pixel values of 8 pixels. The calculation method of the feature amount of other blocks is as described above. An example in which the image 61 is divided into 3 × 3 will be described. It is determined whether or not at least one important pixel is included in each of the nine blocks obtained by the division. Looking at the image 63, all but the three blocks on the right are important blocks. Therefore, since the nine blocks on the right side of the nine blocks of the image 63 are integrated, the seven blocks shown in the image 66 are the final blocks. The integrated block here follows the method for interpreting each block of the image 63 described above, and therefore includes 8 pixels. Therefore, the feature amount of the integrated block is an average of the pixel values of 8 pixels. The calculation method of the feature amount of other blocks is as described above. In this modification, the information indicating the block division method is not only “division method A: {2, 2}” or “division method B: {3, 3}”, but also “region information (C0 to C0). 16 pieces of 1-bit data indicating whether each of C15 is important or not important). Then, information indicating the division method and verification data are output from the image output unit 15 together with the image data I. Thereby, also in the verification apparatus, the block division shown in the images 65 and 66 in FIG. 6 can be performed, and the falsification verification of each block can be normally performed.

  The present invention is also realized by executing the following processing. That is, software (program) for realizing the functions of the above-described embodiments is supplied to a system or apparatus via a network or various storage media, and a computer (or CPU, MPU, etc.) of the system or apparatus reads the program. It is a process to be executed. Therefore, a program that causes a computer to function as the above-described image processing apparatus or verification apparatus by being read and executed by the computer is included in the scope of the present invention. Further, a computer-readable storage medium storing the program is also included in the scope of the present invention.

Claims (8)

An image input unit for inputting image data;
A dividing unit that divides the input image data into a plurality of blocks;
An extraction unit that generates a feature amount for each of the image data divided by the dividing unit;
A generation unit that generates verification data corresponding to the feature amount extracted by the extraction unit;
An image processing apparatus comprising: an output unit that adds information indicating the division method of the division unit and the verification data to the input image data and outputs the added image data. An image input unit for inputting image data;
A dividing unit that divides the input image data into a plurality of blocks;
Generating a feature amount for each of the image data divided by the dividing unit, generating a verification data corresponding to each feature amount; and
An image processing apparatus comprising: an output unit that adds information indicating the division method of the division unit and the verification data to the input image data and outputs the added image data.   The image processing apparatus according to claim 2, wherein the extraction unit generates an average of pixel values constituting the block as a feature amount for each block divided by the division unit.   The image processing apparatus according to claim 2, wherein the verification data is a hash value of the feature amount.   The dividing unit divides the input image data into a plurality of blocks according to a first dividing method, and the generating unit generates a feature amount for each of the image data divided according to the first dividing method, Verification data is generated for each feature amount, the dividing unit divides the input image data into a plurality of blocks according to a second dividing method, and the generating unit divides according to the second dividing method. A feature amount is generated for each of the received image data, and verification data is generated for each of the feature amounts. The output unit includes information indicating the first division method and a first value for the input image data. The image processing apparatus according to claim 2, wherein information indicating two division methods and the verification data are added and output. A method for controlling an image processing apparatus, the image processing apparatus comprising:
A dividing step of dividing the input image data into a plurality of blocks;
Generating a feature amount for each of the divided image data and generating verification data corresponding to each feature amount; and
A method of causing the input image data to execute an output step in which information indicating the division method of the division step and the verification data are added and output.   A program that causes a computer to function as the image processing apparatus according to any one of claims 1 to 5 by being read and executed by a computer.   A computer-readable storage medium storing the program according to claim 7.

Images