JP2019012986A - Verification device, information processing system, verification method, and program - Google Patents

Abstract

To provide a system that can guarantee the authenticity of a group of image information as a whole.SOLUTION: A verification device 11 comprises: accumulation means that accumulates input information including image information, additional information added to the image information, and calculation results of calculations using the image information and a predetermined number of pieces of invariant information; selection means that selects, from the accumulated input information, a predetermined number of pieces of input information; arithmetic means that derives a predetermined number of pieces of invariant information by using the image information and calculation results included in the selected predetermined number of pieces of input information; verification means that compares the invariant information obtained from the additional information with one of the derived predetermined number of pieces of invariant information to verify the authenticity of the predetermined number of pieces of image information included in the selected predetermined number of pieces of input information; determination means that determines the presence/absence of authenticity of the pieces of image information included in the pieces of accumulated input information on the basis of verification results of the verification performed by the verification means; and deletion means that deletes, from the accumulation means, input information including image information determined to be absent of authenticity.SELECTED DRAWING: Figure 5

Description

  The present invention relates to a verification apparatus, an information processing system, a verification method, and a program for causing a computer to execute verification processing for verifying the authenticity of image information.

  An increasing number of moving image data as image information requires authenticity indicating that a moving image such as video from a surveillance camera has not been tampered with. In view of this, a method for guaranteeing the authenticity of moving image data by attaching a signature to the moving image data has been proposed (for example, see Patent Document 1).

  However, in the above method, the moving image data cannot be altered if the moving image data is signed. Therefore, if any one of the group of data is falsified, as a result, there is a problem that the authenticity of all the group of data is lost.

  The present invention has been made in view of the above problems, and an object thereof is to provide an apparatus, a system, a method, and a program that can ensure the authenticity of the entire group of image information.

In order to solve the above-described problem, an embodiment of the invention is a verification device that verifies the authenticity of image information,
Storage means for storing input information including image information, additional information added to a group of image information including image information, and a calculation result calculated using the image information and a predetermined number of invariant information; ,
Selection means for selecting a predetermined number of input information from the input information stored in the storage means;
A calculation means for deriving a predetermined number of invariant information using image information and calculation results included in the predetermined number of input information selected by the selection means;
By comparing the invariant information obtained from the additional information with one of the predetermined number of invariant information derived by the arithmetic means, the predetermined number of image information included in the predetermined number of input information selected by the selection means Verification means for verifying authenticity;
Based on the verification result verified by the verification unit, a determination unit that determines whether or not there is authenticity for each image information included in each input information stored in the storage unit;
There is provided a verification apparatus including a deletion unit that deletes input information including image information determined to have no authenticity by a determination unit from an accumulation unit.

  According to the present invention, the authenticity of the entire group of image information can be ensured.

The figure which showed an example of the system configuration | structure of an information processing system. The figure explaining a secret sharing system. The figure which showed an example of the hardware constitutions of the imaging device which comprises an information processing system. The figure which showed an example of the hardware constitutions of the verification apparatus which comprises an information processing system. The block diagram which showed an example of the function structure of the verification apparatus. The flowchart which showed an example of the process performed by a verification apparatus. The figure explaining the data acquisition timing of a verification apparatus. The figure explaining the 1st process which verifies the signature as additional information. The figure explaining the 2nd process which verifies a signature. The figure explaining the 3rd process which verifies a signature. The figure explaining the 4th process which verifies a signature.

  FIG. 1 shows a system configuration of an information processing system that acquires image information as moving image data, transfers the acquired moving image data with a signature as additional information, and verifies the authenticity of the transferred moving image data. It is the figure which showed an example. The information processing system illustrated in FIG. 1 includes an imaging device 10 that acquires moving image data, transfers it with a signature, and a verification device 11 that verifies the authenticity of the transferred moving image data.

  The imaging device 10 and the verification device 11 are connected by a cable 12. The imaging device 10 and the verification device 11 may be connected wirelessly or via a network such as a LAN (Local Area Network) or a WAN (Wide Area Network).

  The imaging device 10 is a device that continuously captures images and outputs them as moving image data, and transfers the moving image data by dividing it into transmission units such as packets, frames, and segments. In addition to these, the moving image data may be divided into one I frame from a specific I frame.

  Here, when transmitting in units of frames, the data to be transmitted can be compressed by dividing it into I frames, P frames, and B frames. The I frame holds information on all the frames, the P frame holds information on the difference from the previous I frame, and the B frame holds information on the difference between the previous and next I frames and the P frame. is there.

  The verification device 11 uses the moving image data transferred from the imaging device 10 and the signature attached to the moving image data to verify the authenticity of the moving image data by verifying the signature.

  The information processing system performs signature verification when k (k is an integer equal to or less than n) moving image data among n (n is an integer equal to or greater than 1) moving image data can be acquired. A signature method that can be used is adopted. As such a signature scheme, a secret sharing scheme can be adopted.

  Here, the secret sharing scheme will be described with reference to FIG. In the secret sharing scheme, a secret sharing algorithm is used, and Shamir's (k, n) threshold method can be adopted as the secret sharing algorithm. Note that the secret sharing algorithm is not limited to Shamir's (k, n) threshold method, and any algorithm that can receive moving image data and assign it to distributed secret information may be used.

  Shamir's (k, n) threshold method divides secret information into n pieces, and if k of them are used, secret information can be restored, and secret information cannot be restored if k-1 or less. It is. In this method, a k-1th order polynomial f (x) with a constant term as secret information is created, and the solution f (i) when x is i (i is an integer between 1 and n) is divided into n. Assign and distribute to each secret information. Therefore, if k or more i and f (i) can be obtained, f (x) can be obtained by polynomial interpolation, and the constant term f (0) when x = 0, that is, secret information is obtained. Can be identified. On the other hand, if only k-1 or less i and f (i) can be acquired, f (x) cannot be obtained, so that the secret information cannot be specified.

  In the example shown in FIG. 2, a group of moving image data corresponds to secret information, and each moving image data divided by a transmission unit such as a packet corresponds to n pieces of secret information. The k−1th order polynomial f (x) has constants (coefficients) as invariant information in terms of each dimension, and k constants from the constant terms to the k−1th order terms are created. The constant can be a random number generated using, for example, a random number generator. In the following description, random numbers are used as polynomial constants.

The random numbers generated by the random number generator are α _{0 to} α _k−1 . In the example shown in FIG. 2, the constant term is α ₀ . The calculation result (calculated value) Wi of the k−1 degree polynomial can be expressed as the following Expression 1 using a hash function (hash) and moving image data Di. The hash function is a function for obtaining a numerical value representative of the moving image data Di, and the obtained numerical value is a hash value or a hash.

The imaging device 10 includes a random number generator, generates random numbers α _{0 to} α _k−1 , randomly selects them (S1), and assigns them to the constants of the above equation 1. The imaging apparatus 10 calculates n calculated values Wi, that is, W1, W2,..., Wn using Equation 1 to which constants are assigned (S2). The imaging device 10 transfers the moving image data Di and the calculated value Wi calculated from the Di to the verification device 11 as a set of data. A pair of Di and Wi is a pair of data having a property that there is no problem even if some data does not reach the verification device 11. For this reason, the imaging apparatus 10 may ignore some data loss and transfer the data by streaming transfer that emphasizes real-time characteristics. In addition, the imaging apparatus 10 does not have a problem even if some data does not arrive, and therefore does not have a mechanism for retransmitting.

  Note that the calculation of Wi is difficult because the number of bits grows according to the number of k. For this reason, to calculate realistically using a CPU (Central Processing Unit), generate a prime number q with the same number of bits as the hash, and let Wi be the number modulo (divide number) It is desirable.

  The imaging device 10 calculates W1, W2,..., Wn for n pieces of moving image data D1, D2,..., Dn, and then a signature as a signature to be attached to these moving image data D1, D2,. The value S is calculated (S3). Then, the imaging device 10 also transfers the calculated signature value S to the verification device 11. The signature value S can be calculated using a standard signature algorithm (Sigh). For example, an RSA signature, an RSASSA-PSS signature, an ElGamal signature, or the like can be used as a signature scheme algorithm. Since these algorithms are well known, they will not be described in detail here.

The signature value S can be obtained by encoding the constant term α ₀ by the following equation 2 using its own secret key sk _cam held and managed by the imaging apparatus 10.

  Note that when the imaging device 10 transmits the signature value S, if the signature value S does not reach the verification device 11, the authenticity of the moving image data transmitted by the imaging device 10 cannot be guaranteed. The pair of Di and the calculated value Wi of the polynomial is transferred to the verification device 11 (S4). Thereby, the signature value S can be acquired from any data.

  The verification device 11 receives and accumulates the pair of the moving image data Di and the calculated value Wi captured by the imaging device 10 and the signature value S as input information. However, the verification apparatus 11 may have data missing in the Di / Wi pair. This is because there is no retransmission mechanism even if the imaging device 10 does not arrive.

The verification device 11 checks whether or not the number of input information stored in itself is greater than or equal to a predetermined number (k), and if there are k or more, the pair of Di and Wi of the input information is k−1 degree polynomial. The random numbers α _{0 to} α _k−1 generated by the imaging apparatus 10 can be calculated uniquely (S5). This is due to the property that it can be solved when there are k or more pairs of arguments and solutions in an equation that solves a linear equation with k unknowns. If the number of Di / Wi pairs is less than k, the verification apparatus 11 regards the signature verification failure as data that is severely missing and is not reliable (S5).

After calculating the random numbers α _{0 to} α _k−1 , the verification device 11 performs signature verification using a verification (Vrfy) algorithm corresponding to the signature scheme algorithm used in the imaging device 10. Signature verification can be performed by using the public key vk _cam that corresponds to the private key sk _cam of the imaging apparatus 10 and whether or not the following Expression 3 is satisfied.

That is, the signature value S is decrypted using the public key vk _cam, and it is confirmed whether the decrypted value (signed value) matches the constant term α _0. If they do not match, it is considered that the signature verification has failed (S6).

  With reference to FIG. 3, the hardware configuration of the imaging apparatus 10 will be described. The imaging device 10 is a surveillance camera or the like, and includes a lens 20, an image sensor 21, and a sensor controller 22 as in a general camera. The image sensor 21 converts light incident through the lens 20 into an electrical signal and outputs it as image data. As the image sensor 21, for example, an image sensor composed of a CCD (Charge Coupled Device) or a CMOS (Complementary Metal-Oxide Semiconductor) can be used.

  The sensor controller 22 performs exposure control of the image sensor 21, image read control, communication with the verification apparatus 11, image data transmission control, and the like. Here, exposure refers to exposing the image sensor 21 to light incident through the lens 20. The exposure control can be realized, for example, by controlling the exposure time by opening / closing a shutter.

Further, the sensor controller 22 includes a random number generator, generates random numbers α _{0 to} α _k−1, and calculates polynomial calculation values W1 to Wn using moving image data D1 to Dn obtained by imaging. To do. Then, the sensor controller 22 calculates the signature value S and transfers it to the verification device 11 together with the pair of moving image data Di and Wi and the signature value S.

  A hardware configuration of the verification apparatus 11 will be described with reference to FIG. The verification device 11 is a PC, server, tablet terminal, smartphone, MFP (Multi-Function Peripheral), or the like. The verification device 11 includes a CPU 30, a ROM (Read Only Memory) 31, a RAM (Random Access Memory) 32, a HDD (Hard Disk Drive) 33, and a communication I / F 34 as hardware. These are connected to the bus 35 and exchange information with each other via the bus 35.

The CPU 30 controls the entire verification apparatus 11 and executes processes such as storage of input data, calculation of random numbers α _{0 to} α _k−1 , signature verification, and detection of tampered data. The ROM 31 stores a boot program for starting the verification apparatus 11, firmware for controlling the HDD 33, the communication I / F 34, and the like. The RAM 32 provides a work area for the CPU 30.

  The HDD 33 stores an OS (Operating System), input data, a program for executing processing such as calculation of the above random numbers and signature verification. The communication I / F 34 is connected to the imaging device 10, communicates with the imaging device 10, and accepts data input from the imaging device 10.

  The verification device 11 may include other hardware, and may include, for example, an input / output I / F, a display device, an input device, and the like.

  With reference to FIG. 5, the functional configuration of the verification apparatus 11 will be described. The verification device 11 generates a plurality of functional units by the CPU 30 reading and executing a program stored in the HDD 33. Various functions are realized by the plurality of generated functional units. Therefore, it can be said that the verification device 11 includes the plurality of functional units. Here, a plurality of functional units are realized by executing a program, but the present invention is not limited to this, and some or all of the plurality of functional units are realized by hardware such as a dedicated integrated circuit. Also good.

  The verification device 11 includes an input unit 40, a storage unit 41, a selection unit 42, a calculation unit 43, a verification unit 44, a determination unit 45, a control unit 46, and a deletion unit 47 as functional units. The input unit 40 receives input information including the moving image data Di transferred from the imaging device 10, a pair of calculated values Wi for the data, and a signature value S. Then, the input unit 40 stores the received input information in the storage unit 41 and stores it. The input unit 40 stores the input information in the storage unit 41 every time it receives input information, and stores the input information in the storage unit 41.

  The input unit 40 is managed by the control unit 46. In order to verify the signature of the input data, an operation must be performed on k or more pairs for the same signature value S. The control unit 46 determines from which data k or more pairs are to be selected. In addition, when performing streaming transfer, it is not possible to uniquely determine at which timing reception is to be started, so the control unit 46 determines the reception start timing.

  If the input unit 40 starts accepting data input and cannot secure k or more pairs for any signature value S, the control unit 46 determines that the signature verification has failed, and the authenticity cannot be guaranteed. And Then, the control unit 46 sets the data that cannot guarantee the authenticity as the falsified data, and instructs the deletion unit 47 to delete the target data stored in the storage unit 41. The deletion unit 47 deletes the target data from the storage unit 41 so that illegal data does not flow in the subsequent process. For this reason, only the data that can guarantee the authenticity remains in the apparatus, and only the data that can guarantee the authenticity flows to the subsequent process.

  The selection unit 42 operates only when k or more pairs, that is, k or more pieces of input information can be secured for an arbitrary signature value S, and the k or more units stored in the storage unit 41 are operated. Select k pieces of input information from the input information. The selection unit 42 can select k pieces of input information in the order stored in the storage unit 41. As described above, the selection can be made in order from the top, but the selection is not limited to this, and the selection may be made at random.

The calculation unit 43 uses the k pieces of moving image data Di and the k pieces of calculated values Wi included in the k pieces of input information selected by the selection unit 42, for example, a value to which a signature is applied according to the following Expression 4. α ₀ 'is calculated.

The above equation 4 is an equation for obtaining a solution of a k-1 order polynomial by Lagrange interpolation. Lagrangian interpolation is an interpolation method in which a function passing through all k discrete points is obtained and a value is estimated. In this method, the value of α ₀ ′ can be uniquely determined for k or more points. However, in the case of less than k pairs, the same value as α ₀ ′ is not obtained except for an event that can be ignored stochastically. Since any pair can be used as long as there are k pairs, there is no need to forcibly select k pairs.

  When verifying the signature, the verification unit 44 extracts the signature value S from the input information stored in the storage unit 41 and holds it. Since the n pieces of input information including the moving image data D1 to Dn to be verified have the same signature, there is only one signature. Therefore, when there are two or more types of signature values in the signature verification result, the verification unit 44 always fails the signature verification.

When the signature value S of n pieces of input information subject to signature verification is one, the verification unit 44 applies to the following equation 5 and applies the signature value S to the verification algorithm, and the obtained value is calculated by the above equation 4. Confirm that it matches α ₀ '. The verification unit 44 determines that the signature verification is successful if they match, and determines that the signature verification fails if they do not match.

  If the signature verification fails, the verification unit 44 requests the deletion unit 47 to delete the target data. The deletion unit 47 deletes the target data from the storage unit 41, but the target data can also be left for management as an error log in another storage unit or the like.

The selection unit 42 can select at least one of the k pieces of input information stored in the storage unit 41 by replacing at least one of the previously selected pieces of input information, and selecting k pieces of input information again. In response to this, the calculation unit 43 uses the k moving image data Di and the k calculated values Wi included in the k pieces of input information selected by the selection unit 42 to apply a signature from Equation 4 above. The calculated value α ₀ ′ can be calculated. Then, the verification unit 44 can apply to the above formula 5, apply the signature value S to the verification algorithm, and confirm whether the obtained value matches α ₀ ′ calculated by the above formula 4. The verification unit 44 determines that the signature verification is successful if they match, and determines that the signature verification fails if they do not match. This can be repeated until all input information having the same signature value S stored in the storage unit 41 is selected.

  The determination unit 45 determines whether each moving image data included in each input information has authenticity based on the verification result selected by the selection unit 42 at least once and verified by the verification unit 44. The verification unit 44 verifies the authenticity of the k moving image data as a lump. For this reason, when k pieces of moving image data are stored in the storage unit 41 and the k pieces of moving image data are determined to be authentic, each moving image data is determined to be authentic in one selection. Can be determined. On the other hand, if there are more than k pieces of moving image data stored in the storage unit 41, or if it is determined that the moving image data is not authentic even if there are k pieces of stored moving image data, the selecting unit The authenticity is verified by changing the combination according to 42. This makes it possible to determine which moving image data has authenticity or not.

  The deletion unit 47 deletes the input information including the moving image data determined as having no authenticity by the determination unit 45 from the storage unit 41. As a result, only the data whose authenticity can be ensured remains in the apparatus, and is distributed from the storage unit 41 to other devices as a post-process, where data display, data storage, and the like are performed.

  The flow of processing executed by the verification apparatus 11 shown in FIG. 5 is summarized in FIG. In step 605, the input unit 40 starts data input, and the storage unit 41 stores the signature value S and input information including a pair of moving image data and a calculated value. In step 610, it is determined whether the control unit 46 has secured k or more pieces of input information for an arbitrary signature value S. If secured, the process proceeds to step 615; otherwise, the control unit 46 instructs the deletion unit 47 to delete the input information, assuming that the moving image data included in the secured input information is not authentic. The process proceeds to step 655.

In step 615, k or more pieces of input information can be secured for an arbitrary signature value S, and k or more pieces of input information are stored in the storage unit 41. Select k pieces of input information from. In step 620, the calculation unit 43 acquires k moving image data and k calculated values included in the k pieces of input information selected by the selection unit 42, and applies them to the above equation 4 using them. Then, α ₀ ′ is calculated.

In step 625, the verification unit 44 extracts the signature value S from the input information stored in the storage unit 41, applies it to Equation 5 above, applies the signature value S to the verification algorithm, and the obtained value is α ₀ ′. Check if they match. If they match, the process proceeds to step 630 and the signature verification succeeds. If they do not match, the process proceeds to step 635 and the signature verification fails.

  In step 640, the determination unit 45 determines whether or not there is authenticity for each of k pieces of moving image data included in the k pieces of input information selected by the selection unit 42 based on the verification result. In step 645, it is confirmed whether or not there is moving image data that has not been determined for authenticity among k or more moving image data included in the k or more input information accumulated in the accumulating unit 41. If there is moving image data that has not yet been determined, the process returns to step 615, and the selection unit 42 changes the combination and selects it again for verification.

  On the other hand, if there is no moving image data that has not been determined, the process proceeds to step 650 to check whether there is moving image data that has been determined not to be authentic. If there is, the process proceeds to step 655, and the deletion unit 47 receives input information including moving image data determined to have no authenticity from the storage unit 41 in response to an instruction from the control unit 46 or a determination result from the determination unit 45. delete. Then, the process proceeds to Step 660. On the other hand, if there is no moving image data determined as having no authenticity at step 650, the process proceeds directly to step 660.

  In step 660, the control unit 46 confirms whether there is data input of another signature value S. If yes, return to Step 610; otherwise, go to Step 665 and end the process.

  With reference to FIG. 7, the data acquisition timing determined by the control unit 46 when performing streaming reception will be described. In order to facilitate understanding, a specific example will be described. Here, n = 5, k = 3, moving image data is a frame, and there are five frames having the same signature value S. If three or more frames can be acquired, verification can be performed. To do.

  The control unit 46 manages data input by the input unit 40, instructs the input unit 40, and starts frame input. In the example shown in FIG. 6, two frames can be obtained for the signature value A, but the condition that k is 3 or more is not satisfied. Therefore, the control unit 46 instructs the deletion unit 47 to delete the frame having the signature value A, and the deletion unit 47 deletes the frame having the signature value A stored in the storage unit 41. The storage unit 41 stores a signature value A, a pair of a frame and a calculated value as input information, and the deletion unit 47 deletes the signature value A, the frame, and the calculated value.

Next, the signature value is switched to B, and five frames can be obtained for the frame having the signature value B, and k satisfies the condition of 3 or more. For this reason, the selector 42 is instructed to execute the next process. Here, the selection unit 42 selects three frames, the calculation unit 43 calculates α ₀ ′, the verification unit 44 applies the signature value B to the verification algorithm, and the obtained value and the α ₀ calculated by the calculation unit 43 are calculated. Compare with 'and perform signature verification. Then, the same processing is repeated by changing the frame selected by the selection unit 42, and the determination unit 45 determines which of the five frames is authentic or not based on the obtained verification result.

  In this way, frame input may be started and signature verification may be performed from the signature value attached to the acquired frame, but the next signature value is not verified for the frame having the signature value acquired first. Signature verification may be performed from a frame having This method is based on the premise that acceptance of input starts in the middle of a frame, and a frame having the first signature value is discarded. Therefore, in the example shown in FIG. 7, it is possible to skip the frame having the signature value A, accept input from the frame having the signature value B, and perform signature verification.

  The determination process performed by the determination unit 45 will be described with reference to FIGS. FIG. 8 is a diagram illustrating a first example of determination processing performed by the determination unit 45. Also in this example, it is assumed that n = 5 and k = 3, and there are five frames having the same signature value S, and if three or more frames can be acquired, verification can be performed.

  In the example shown in FIG. 8, even if one (or two) frames out of the five frames having the signature value A are missing, if three frames can be acquired, the presence / absence of authenticity is determined. The target data. For the signature value A, the selection unit 42 selects the first three frames (A-1, A-2, A-3), the calculation unit 43 performs the calculation, and the verification unit 44 includes these three frames. Verify the authenticity of the set of frames.

  In the example shown in FIG. 8, since only the frame A-4 is falsified data, the verification unit 44 verifies that the signature verification is successful for the frames A-1, A-2, and A-3. Output the result. The determination unit 45 determines that the frames A-1, A-2, and A-3 have authenticity in response to the output verification result.

  Next, the selection unit 42 changes the frame A-3 to the frame A-4, for example, and selects the frames A-1, A-2, and A-4. The calculation unit 43 performs calculations using the selected frames A-1, A-2, and A-4, and the verification unit 44 verifies the authenticity of the frame group composed of these three frames. In this case, since the frame A-4 that is falsified data is included, the verification unit 44 outputs a verification result indicating that the signature verification failed.

  The determination unit 45 determines that the frames A-1, A-2, and A-3 are authentic, and the verification unit 44 performs the frame group of the frames A-1, A-2, and A-4. Since the verification result is a signature verification failure, it is determined that the frame A-4 is falsified data and the frame A-4 is not authentic.

  Then, the selection unit 42 changes the frame A-4 to the frame A-5, for example, and selects the frames A-1, A-2, and A-5. The calculation unit 43 performs calculations using the selected frames A-1, A-2, and A-5, and the verification unit 44 verifies the authenticity of the frame group composed of these three frames. In this case, the verification unit 44 outputs a verification result indicating that the signature verification is successful. The determination unit 45 determines that the frames A-1 and A-2 are authentic, and the verification result obtained by the verification unit 44 for the frames A-1, A-2, and A-5 is a signature. Since the verification is successful, the frame A-5 is also determined as data having authenticity.

  Note that the determination unit 45 changes the frame instead of every time the selection unit 42 selects a frame, and whether or not there is authenticity for the frames A-1 to A-5 based on all verified verification results. May be determined.

  The determination unit 45 instructs the deletion unit 47 to delete the frame A-4 determined to have no authenticity. In response to the instruction from the determination unit 45, the deletion unit 47 deletes the instructed frame A-4 from the storage unit 41.

  In this way, only the falsified data is deleted from the storage unit 41, and the data that has not been falsified remains in the storage unit 41. Therefore, the authenticity of the entire moving image data having the same signature value can be ensured. .

  FIG. 9 is a diagram illustrating a second example of the determination process performed by the determination unit 45. In this example, if n ≧ 2k, for example, n = 10, k = 3, and the number of frames having the same signature value S is 10, and if three or more of the frames can be acquired, verification can be performed. To do.

  For ten frames having a signature value X, three frames can be grouped as a set of frames to form three groups. For example, a group of frames composed of frames X-1, X-2, and X-3 is group K-1, and a group of frames composed of frames X-4, X-5, and X-6 is group K-2. And a group of frames composed of frames X-7, X-8, and X-9 can be group K-3.

  The selection unit 42 selects three frames that constitute one of the groups K-1, K-2, and K-3. Here, the selection unit 42 selects the frames X-1, X-2, and X-3 that constitute the group K-1. The calculation unit 43 performs calculation using the selected frames X-1, X-2, and X-3, and the verification unit 44 verifies the authenticity of the group K-1 composed of these three frames. In this case, since the frame X-9 that is falsified data is not included, the verification unit 44 outputs a verification result that the signature verification is successful. The determination unit 45 receives the verification result that the signature verification is successful from the verification unit 44, and determines that the frames X-1, X-2, and X-3 constituting the group K-1 are authentic.

  Similarly, the selection unit 42 selects three frames X-4, X-5, and X-6 constituting the group K-2, the calculation unit 43 performs the calculation, and the verification unit 44 performs the group K-2. Verify the authenticity of. Also in this case, the verification unit 44 outputs a verification result that the signature verification is successful. The determination unit 45 receives the verification result that the signature verification is successful from the verification unit 44, and determines that the frames X-4, X-5, and X-6 constituting the group K-2 are authentic.

  Similarly, the selection unit 42 selects three frames X-7, X-8, and X-9 constituting the group K-3, the calculation unit 43 performs the calculation, and the verification unit 44 performs the group K-3. Verify the authenticity of. In this case, since the verification unit 44 includes the frame X-9 that is falsified data, the verification unit 44 outputs a verification result indicating that the signature verification failed.

  In order to verify in detail the group K-3 of the failed part, the selection unit 42, for example, a part of the frame constituting the group K-1 in which the signature verification was successful and one of the frames constituting the failed group K-3. Select one. Specifically, for example, the selection unit 42 selects frames X-1 and X-2 from the group K-1, and selects a frame X-7 from the group K-3. This is to determine whether or not the frame X-7 is authentic. Note that the frame selected from the group K-1 may be X-1, X-3, or X-2, X-3. Also, two frames in group K-2 may be used.

  The calculation unit 43 performs calculation using the frames X-1, X-2, and X-7 selected by the selection unit 42, and the verification unit 44 verifies the authenticity of the frame group composed of these three frames. To do. In the example shown in FIG. 8, since none of the data is falsified, the verification unit 44 outputs a verification result indicating that the signature verification is successful. In response to this, the determination unit 45 determines that the frame X-7 is authentic.

  Next, the selection unit 42 selects the frames X-1 and X-2 from the group K-1, and selects the frame X-8 from the group K-3. This is to determine the authenticity of frame X-8. Similarly in this case, the frame selected from the group K-1 may be X-1, X-3, or X-2, X-3. Also, two frames in group K-2 may be used.

  The calculation unit 43 performs calculation using the frames X-1, X-2, and X-8 selected by the selection unit 42, and the verification unit 44 determines the authenticity of the frame group composed of these three frames. Validate. In the example shown in FIG. 8, since none of the data is falsified, the verification unit 44 outputs a verification result indicating that the signature verification is successful. In response to this, the determination unit 45 determines that the frame X-8 is authentic.

  The determination unit 45 determines the verification result of the signature verification failure for the group K-3 output from the verification unit 44 and the data having authenticity for the frames X-7 and X-8 constituting the group K-3. Therefore, it can be determined that the remaining frames X-9 constituting the group K-3 are not authentic.

  Similarly to the above, the selection unit 42 selects the frames X-1, X-2, and X-9, receives the verification result of the signature verification failure from the verification unit 44, and the determination unit 45 receives the frame X Since -1 and X-2 have authenticity, it may be determined that frame X-9 has no authenticity.

  For frame X-10 that does not belong to the group, it is possible to determine whether it is authentic by performing verification using a part of the frames that make up the groups K-1 and K-2 that have been successfully verified. . Specifically, the selection unit 42 selects, for example, X-1 and X-2 and the frame X-10 constituting the group K-1, the calculation unit 43 performs calculation, and the verification unit 44 performs these three operations. The authenticity of a frame group composed of two frames is verified. In the example shown in FIG. 8, since the frame X-10 is data that has not been tampered with, the verification unit 44 outputs a verification result indicating that the signature verification was successful. In response to this, the determination unit 45 determines that the frame X-10 is authentic.

  The determination unit 45 instructs the deletion unit 47 to delete the frame X-9 determined to have no authenticity. The deletion unit 47 receives the instruction from the determination unit 45 and deletes the instructed frame X-9 from the storage unit 41.

  By dividing into groups and verifying for each group, the selection unit 42 selects, the calculation unit 43 performs calculation, and the verification unit 44 reduces the number of verifications, thereby improving the authenticity data extraction efficiency. it can.

  FIG. 10 is a diagram illustrating a third example of the determination process performed by the determination unit 45. In this example, if n <2k, for example, n = 10, k = 6, 10 frames with the same signature value S, and if more than 6 frames can be obtained, verification can be performed. To do.

  In this example, even if grouping is performed, only one group can be created, so grouping is not performed. The selection unit 42 selects a frame group L-1 composed of six frames from the top to the sixth frame, the calculation unit 43 performs a calculation, and the verification unit 44 verifies the authenticity of the frame group L-1. To do.

  In the example shown in FIG. 9, since none of the frames constituting the frame group L-1 is falsified data, the verification unit 44 outputs a verification result indicating that the signature verification is successful. In response to this, the determination unit 45 determines that any frame constituting the data group L-1 is authentic.

  The selection unit 42 replaces the frame group M-1 ′ composed of four frames out of the six frames previously verified with the frame group M-1 composed of the remaining four frames, Two frames not included in the data group M-1 ′ and four frames constituting the M-1 are selected. The calculation unit 43 performs calculation using the six frames selected by the selection unit 42, and the verification unit 44 verifies the authenticity of the frame group composed of these six frames.

  When the verification unit 44 outputs a verification result indicating that the signature verification is successful, the determination unit 45 determines that the four frames constituting the data group M-1 are also authentic. On the other hand, when the verification unit 44 outputs a verification result indicating that the signature verification has failed, the four frames constituting the data group M-1 are further divided into two, and four of the six frames that have been verified first. The authenticity of the frame group composed of the frame and one of the two divided frames is verified.

  Here, if the signature verification is successful, it is determined that the two frames are authentic. Similarly, the authenticity of the remaining two frames can be verified. On the other hand, if it is determined that there is no authenticity, either one or both of the two frames are not authentic. Therefore, the verification is performed first, and five of the six frames in which the signature verification is successful The authenticity of each of the two frames is verified using one of the two frames.

  FIG. 11 is a diagram illustrating a fourth example of the determination process performed by the determination unit 45. In this example, falsified data exists in a group of frames having the same signature value X. In this case, even if verification is performed on the frame group L-1, signature verification fails, and verification is performed by replacing the frame group M-1 ′ in the frame group L-1 with the remaining frame group M-1. , Signature verification fails. If the tampered data is scattered and the number is large, the number of combinations to be selected by the selection unit 42 increases, and verification takes time. In the verification, a processing time may be set as a constraint condition such as a constraint item.

  Therefore, for a frame for which verification is not completed within the processing time, the frame can be excluded from the target. Therefore, the frame can be deleted from the storage unit 41. In the worst case, all the frames having the signature value X are deleted from the storage unit 41.

  This is because if a group of frames including data that has been tampered with before the verification is completed is accumulated and provided, the authenticity of the entire moving image data cannot be guaranteed. Therefore, in the present system, in order to give first priority to guaranteeing authenticity, a frame or a frame group that may contain tampered data is deleted when it is not possible to determine whether or not authenticity exists.

  The present invention has been described with the embodiments described above as a verification apparatus, an information processing system, a verification method, and a program. However, the present invention is not limited to the above-described embodiments, and other embodiments, additions, modifications, deletions, and the like can be modified within a range that can be conceived by those skilled in the art. . For example, when it is only necessary to extract minutes or seconds of data in a moving image as evidence, such as a crime scene, it has been difficult to extract only the necessary data by attaching a signature. However, according to the present invention, it is possible to flexibly cope with such a situation. In addition, any aspect is included in the scope of the present invention as long as the operations and effects of the present invention are exhibited. Therefore, a recording medium on which the program is recorded, a program providing server that provides the program, and the like can also be provided.

DESCRIPTION OF SYMBOLS 10 ... Imaging device 11 ... Verification device 12 ... Cable 20 ... Lens 21 ... Image sensor 22 ... Sensor controller 30 ... CPU
31 ... ROM
32 ... RAM
33 ... HDD
34 ... Communication I / F
35 ... Bus 40 ... Input unit 41 ... Storage unit 42 ... Selection unit 43 ... Calculation unit 44 ... Verification unit 45 ... Determination unit 46 ... Control unit 47 ... Deletion unit

Japanese Patent No. 5151987

Claims (9)

A verification device for verifying the authenticity of image information,
Accumulating input information including the image information, additional information added to a group of image information including the image information, and a calculation result calculated using the image information and a predetermined number of invariant information Storage means;
Selection means for selecting a predetermined number of input information from the input information stored in the storage means;
Computing means for deriving the predetermined number of invariant information using image information and calculation results included in the predetermined number of input information selected by the selection means;
By comparing the invariant information obtained from the additional information with one of the predetermined number of invariant information derived by the arithmetic means, a predetermined number of input information selected by the selection means includes Verification means for verifying authenticity of image information;
Based on a verification result verified by the verification unit, a determination unit that determines whether or not there is authenticity for each image information included in each input information stored in the storage unit;
A verification apparatus comprising: deletion means for deleting input information including image information determined to be unauthenticated by the determination means from the storage means.   Determining whether or not the number of the input information is equal to or greater than a predetermined number, and including control means for instructing the deletion means to delete the input information from the storage means when the number is less than the predetermined number; The verification device according to claim 1. The selection unit replaces at least one of a predetermined number of pieces of input information in which the image information is verified as authentic by the verification unit,
The determination unit determines whether or not each image information is authentic based on a verification result for the predetermined number of image information and a verification result exchanged by the selection unit and verified by the verification unit. The verification apparatus according to 1 or 2. When the number of pieces of image information stored in the storage unit is more than twice a predetermined number, the selection unit creates two or more sets including a predetermined number of pieces of image information. Select one,
The verification device according to claim 1, wherein the verification unit verifies the authenticity of a predetermined number of pieces of image information constituting each set selected by the selection unit. When the number of image information stored in the storage unit is less than twice the predetermined number, the selection unit selects a part of the predetermined number of input information verified by the verification unit as being authentic. , Replacing all or part of the input information including image information that has not been determined whether there is authenticity stored in the storage means,
The determination unit determines whether or not each image information is authentic based on a verification result for the predetermined number of image information and a verification result exchanged by the selection unit and verified by the verification unit. The verification apparatus of any one of 1-3.   The said determination means instruct | indicates to delete the input information containing the image information which has not determined the presence or absence of authenticity with respect to the said deletion means based on constraint conditions. The verification device described. An input that captures an image and includes image information, additional information added to a group of image information including the image information, and a calculation result calculated using the image information and a predetermined number of invariant information. An information processing system comprising: an imaging device that transmits information; and a verification device that receives the input information from the imaging device and verifies the authenticity of the image information,
The verification device is
Storage means for storing the input information;
Selection means for selecting a predetermined number of input information from the input information stored in the storage means;
Computing means for deriving the predetermined number of invariant information using image information and calculation results included in the predetermined number of input information selected by the selection means;
By comparing the invariant information obtained from the additional information with one of the predetermined number of invariant information derived by the arithmetic means, a predetermined number of input information selected by the selection means includes Verification means for verifying authenticity of image information;
Based on a verification result verified by the verification unit, a determination unit that determines whether or not there is authenticity for each image information included in each input information stored in the storage unit;
An information processing system comprising: deletion means for deleting, from the storage means, input information including image information determined to be non-authentic by the determination means. A verification method for verifying the authenticity of image information,
Storage means for storing input information including the image information, additional information added to a group of image information including the image information, and a calculation result calculated using the image information and a predetermined number of invariant information The steps to accumulate in
Selecting a predetermined number of input information from the input information stored in the storage means;
Deriving the predetermined number of invariant information using image information and calculation results included in the selected predetermined number of input information;
The authenticity of a predetermined number of pieces of image information included in the selected predetermined number of input information is verified by comparing the invariant information obtained from the additional information with one of the derived predetermined number of invariant information. Steps,
Based on the verification result verified in the verification step, determining whether or not there is authenticity for each image information included in each input information stored in the storage unit;
And a step of deleting, from the storage means, input information including image information determined not to be authentic in the determining step.   A program for causing a computer to execute each step included in the verification method according to claim 8.