KR102101557B1 - Image authentication method and apparatus based on object detection - Google Patents

Abstract

Disclosed is an object recognition-based image authentication method and apparatus. The image authentication method includes: (a) extracting each object from the image; (b) encrypting each extracted object with a different key; And (c) generating a digital signature for each of the encrypted objects and the background image using a secret key at a specific point in time.

Description

Image authentication method and apparatus based on object detection}

The present invention relates to an object recognition-based image authentication method and apparatus.

With the advancement of technology, numerous control devices (CCTV, black box, etc.) are installed in the element element. As these control devices increase, videos used as evidence in the event of a crime are also increasing.

However, as the image photographed by the control device is used as evidence as it is, the problem that the privacy of a victim or a third party is unintentionally infringed occurs very frequently. To prevent this, the victim or a third party may be masked in the video. However, in order for the image to be used as evidence, it must be able to guarantee that the image is an original. However, if the image is tampered with masking, there is a problem in that it cannot be guaranteed because the digital signature inserted into the image is invalid.

The present invention is to provide an object recognition-based image authentication method and apparatus capable of protecting an individual's privacy by encrypting each object with a different key and selectively decrypting only a specific object.

According to an aspect of the present invention, there is provided an image authentication method capable of preventing an invasion of personal privacy in an image.

According to an embodiment of the present invention, (a) extracting each object from the image, respectively; (b) encrypting each extracted object with a different key; And (c) generating a digital signature for each of the encrypted objects and the background image using a secret key at a specific point in time.

The step (b) includes: extracting an encryption key for each object from the master key; And encrypting each object with a different key using the extracted encryption key for each object.

The background image is a region other than each extracted object from the image.

After generating the digital signature, the method may further include updating the secret key for digital signature at another time.

The digital signature can be generated using a forward secure signature algorithm.

According to another embodiment of the present invention, generating a decryption key for an object to be decoded from an encrypted image-the encrypted image includes a background image and an encrypted object, and includes a digital signature; Decoding the object to be decoded using the decoding key; And combining and outputting the decoded object to be decoded to the background image.

Before the step of generating the decryption key, the encrypted image is digitally signed with a secret key at a specific point in time, and may further include verifying the digital signature with a public key corresponding to the secret key.

According to another aspect of the present invention, an apparatus capable of preventing an invasion of personal privacy in an image is provided.

According to an embodiment of the present invention, an image authentication apparatus comprising: an object extraction unit for extracting each object from an image; An encryption unit that encrypts each extracted object with a different key; And a signature generator that generates a digital signature for each of the encrypted objects and background images using a secret key at a specific time.

Further comprising a key generator for generating keys for object encryption and digital signature, the key generator is used to generate a master key used for encrypting each object and a secret key and a public key pair for digital signature at a specific point in time. You can.

The encryption unit may extract an encryption key for each object from the master key, and then encrypt each object with a different key using each encryption key extracted for each object.

After generating the digital signature, an update unit for updating the secret key for digital signature at another time may be further included.

According to another embodiment of the present invention, a key generation unit for generating a decryption key for an object to be decoded from an encrypted image-the encrypted image includes a background image and an encrypted object, and includes a digital signature; A decoding unit decoding the object to be decoded using the decoding key; And an output unit that combines and outputs the decoded object to be decoded to the background image.

By providing an object recognition-based image authentication method and apparatus according to an embodiment of the present invention, each object is encrypted with a different key, and there is an advantage of protecting an individual's privacy by selectively decoding only a specific object.

1 is a diagram schematically showing a system configuration according to an embodiment of the present invention.
2 is a diagram illustrating a video authentication process according to an embodiment of the present invention.
Figure 3 is a block diagram showing the internal configuration of the video authentication device included in the control device according to an embodiment of the present invention.
Figure 4 is a flow chart showing a video authentication method according to an embodiment of the present invention.
5 is a diagram illustrating a pseudo code for key generation according to an embodiment of the present invention.
6 is a diagram illustrating a pseudo code for updating a secret key according to an embodiment of the present invention.
7 is a diagram illustrating a pseudo code for encrypting each object by using an encryption key for each object according to an embodiment of the present invention, and generating a digital signature for the background image and the encrypted object.
8 is a flowchart illustrating an image processing method according to an embodiment of the present invention.
9 is a view schematically showing the internal configuration of an image processing apparatus according to an embodiment of the present invention.
10 is a diagram illustrating a pseudo code for selectively decoding only a specific object according to an embodiment of the present invention and combining it with a background image for output.

The singular expression used in this specification includes the plural expression unless the context clearly indicates otherwise. In this specification, terms such as “consisting of” or “comprising” should not be construed as including all of the various components, or various steps described in the specification, among which some components or some steps are It may not be included, or it should be construed to further include additional components or steps. In addition, terms such as “... unit” and “module” described in the specification mean a unit that processes at least one function or operation, which may be implemented in hardware or software, or a combination of hardware and software. .

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram schematically showing a system configuration according to an embodiment of the present invention, and FIG. 2 is a diagram illustrating a video authentication process according to an embodiment of the present invention.

Referring to FIG. 1, a system according to an embodiment of the present invention includes a control device 110 and an image processing device 120.

The control device 110 is a device that photographs images in real time, and may be, for example, a CCTV or a black box. The control device 110 does not just record a video in real time, but includes a video authentication device 300, and recognizes each object for each frame of a video through the video authentication device, and then encrypts each object. You can create a digital signature. The video authentication device 300 may sign the encrypted object based on a forward secure signature scheme. This will be described in more detail with reference to FIG. 2 below.

The image processing apparatus 120 selectively decodes only the object to be decoded using the decoding key of the object to be decoded. Since the image is captured by the control device 110 and each object in the image is encrypted and stored with a different key, when viewing the image, it appears to be masked. Accordingly, the image processing apparatus 120 may selectively decode only the specific object based on the key for the object to be decoded so that only a specific object in the image can be selectively viewed for a specific purpose.

The process of selectively decoding each object in an image according to an embodiment of the present invention by encrypting and signing with a different key will be described with reference to FIG. 2.

It is assumed that the original image is photographed in real time as shown in (a) of FIG. 2. In the original image, each object is recognized as shown in Fig. 2B. Although (b) of FIG. 2 is shown to recognize only the 210a and 210b objects, it is natural that the recognized objects may vary depending on the implementation method. Preferably, all objects included in the image can be recognized.

According to an embodiment of the present invention, when an image is used for a specific purpose (eg, criminal investigation, court filing, etc.), for the purpose of preventing the faces of people other than the person actually involved from being indiscriminately exposed As implemented, it is assumed that people are recognized as individual objects. However, it is natural that other objects may be recognized. In addition, although not described in detail herein, object recognition may use a deep learning technique. The method of recognizing each object in an image (frame) based on deep learning technology is obvious to a person skilled in the art and is not related to the main issue of the present invention, so a detailed description thereof will be omitted.

As described above, after each object in the image (frame) is recognized, as illustrated in FIG. 2C, each recognized object may be extracted. Subsequently, as illustrated in (d) of FIG. 2, the stream of each extracted object is compressed using a known video compression technique. Accordingly, as shown in FIG. 2D, a plurality of object video streams can be generated. And, as shown in (e) of FIG. 2, each object is encrypted using each key, and each encrypted object can be digitally signed with the background. As already described above, the digital signature is signed using the omni-directional safety signature technique, and there is an advantage in that it is impossible to operate after a certain point in time due to the characteristics of the omni-directional safety signature technique. The forward safety signature will be described in more detail below.

When the encrypted object frame is opened, a decryption key is issued, and as shown in FIG. 2 (f), a decryption process may be performed after verifying the integrity of the encrypted frame. At this time, by selectively decoding only a specific object using only the decoding key for the specific object to be decoded, as shown in FIG. 2 (f), other objects are not decoded and appear to be masked. It has the advantage of protecting personal privacy. It will be described in more detail below.

3 is a block diagram showing the internal configuration of a video authentication device included in the control device according to an embodiment of the present invention. Here, the video authentication device 300 may be a component of a control device such as a CCTV or a black box or a separate device interlocked with the corresponding control device.

Referring to FIG. 3, the image authentication device 300 according to an embodiment of the present invention includes an image acquisition unit 310, an object extraction unit 315, an encryption unit 320, a signature generation unit 325, and a memory ( 330) and a processor 335.

The image acquisition unit 310 acquires an image captured in real time.

For example, the image acquisition unit 310 may acquire an image captured through a camera or image sensor.

The object extraction unit 315 is a means for recognizing and extracting objects in an image. For example, the object extraction unit 315 may detect an object in an image using known SVM, Haar feature point, HOG, LBP feature point, etc., and recognize and extract an object in the image using a deep learning technique. It might be. Of course, it is natural that the object extraction unit 315 can recognize and extract objects in an image through various object recognition methods in addition to the above-described methods. The object region extracted from the image may be expressed as a null value, as shown in FIG. 2B, and the extracted object region is shown as being masked.

The encryption unit 320 encrypts each object extracted by the object extraction unit 315.

At this time, the encryption unit 320 may encrypt each object using a different encryption key for each object. For example, the encryption unit 320 may encrypt each object by applying a different encryption key for each object based on the master key. Here, the master key may be input separately by the key manager, and it is natural that the corresponding image authentication device 300 may generate it.

The signature generator 325 generates signatures for the remaining images (hereinafter referred to as background images) except for the encrypted object and the extracted object, respectively. The signature generator 325 according to an embodiment of the present invention may generate a signature for an encrypted object and a background image, respectively, using, for example, a forward secure signature algorithm.

In order to facilitate understanding and explanation, the forward secure signature algorithm will be briefly described.

)으로 구성된다. The omni-directional safety signature scheme has four algorithms (

).

는 보안 파라미터(n)와 전체 시간 주기(T)를 입력하는 경우,

키 페어를 출력한다.

When entering the security parameter (n) and the total time period (T),

Print the key pair.

는 현재 시점에 대한 비밀키(

)에 상응하여 다음 시점을 위한 다른 비밀키(

)를 출력한다.

Is the secret key for the current point (

) Corresponding to another secret key for the next point (

).

는 메시지(m), 현재 시점(t)의 비밀키(

)를 입력하는 경우 해당 메시지(m)에 대한 서명을 출력한다.

Is the message (m), the secret key of the current time (t) (

), The signature for the message (m) is output.

는 메시지(m)에 대한 서명이 공개키(

)로 검증되는 경우 제1 값(예를 들어, 1)을 출력하고, 그렇지 않으면 제2 값(예를 들어, 0)을 출력한다.

Is the public key (

), The first value (for example, 1) is output, and the second value (for example, 0) is output.

The signature generation unit 325 may respectively generate an object signature using a secret key for a specific time point for each object detected in the image based on the Forward secure signature scheme described above. As described above, in the forward secure signature scheme, after the secret key for a specific time is generated, the corresponding secret key is updated to generate a signature for an encrypted object and a background image at different times, respectively.

Therefore, even if the secret key is illegally leaked, it is practically impossible to forge an object signature due to a different point in time. Accordingly, in an embodiment of the present invention, a signature for each encrypted object and background image may be generated based on the Forward secure signature scheme.

The memory 330 is a means for storing various algorithms required to perform an image authentication method according to an embodiment of the present invention and various data derived from this process.

The processor 335 includes internal components (eg, an image acquisition unit 310, an object extraction unit 315, an encryption unit 320, and a signature generation unit) of an image authentication device according to an embodiment of the present invention ( 325) and the memory 330).

4 is a flowchart illustrating an image authentication method according to an embodiment of the present invention, FIG. 5 illustrates a pseudo code for key generation according to an embodiment of the present invention, and FIG. 6 illustrates an embodiment of the present invention. A pseudo code for updating a secret key is illustrated, and FIG. 7 encrypts each object using an encryption key for each object according to an embodiment of the present invention, and generates a digital signature for the background image and the encrypted object The pseudo code is illustrated.

In step 410, the image authentication device 300 recognizes and extracts at least one object in the image. Various methods for detecting an object in an image are known, and since the object detection itself is independent of the main issue of the present invention, a detailed description thereof will be omitted. That is, various known methods for detecting an object in an image can be applied without limitation.

In step 415, the image authentication device 300 encrypts each object with a different encryption key and generates three keys necessary to generate a digital signature.

Figure 5 shows the pseudo code for key generation. This will be briefly described with reference to FIG. 5.

)와 특정 시점(T)를 입력받는다. In order to generate a key, the video authentication device 300 may include security parameters (

) And a specific point in time (T).

는 보안 파라미터(

)를 이용하여 암호화를 위한 마스터 키(

)를 생성한다.

Is the security parameter (

) To encrypt the master key (

).

는 보안 파라미터(

)와 특정 시점(T)을 이용하여 디지털 서명을 위한 공개키 및 비밀키의 키 페어

를 생성한다.

Is the security parameter (

) And key pair of public key and secret key for digital signature using specific point in time (T)

Produces

In FIG. 4, the image authentication device 300 is described as generating a key pair of a public key and a secret key for digital signature and a master key for encryption of each object, but the key generation is not the video authentication device 300. Naturally, it may be performed by a separate device. That is, three keys may be generated by a separate key management device (not shown) and then input to the image authentication device 300.

In step 420, the image authentication device 300 updates the secret key.

The pseudo code for this is as shown in FIG. 6. As described above, each encrypted object and background image is digitally signed using a respective omni-directional safety signature algorithm. Such an omni-directional safety signature algorithm can update the secret key to generate a unique signature at another time after using the secret key at a specific time. The pseudo code for this is shown in FIG. 6.

In step 425, the video authentication device 300 generates an encryption key for each object using the master key. Here, the encryption key generated for each object is generated differently according to each object.

In step 430, the image authentication device 300 encrypts each object using the encryption key generated for each object.

In step 435, the image authentication device 300 generates a digital signature for the encrypted object and the background image using a secret key at a specific time.

The background image is an image including the remaining region excluded (masked) from the extracted object region. For example, it will be described with reference to Fig. 2 (c).

As shown in (c) of FIG. 2, when the person1 and person2 objects are extracted from the original image, the remaining areas exist as shown in 220. In the background image, each extracted object area may appear to be masked as no information exists.

The pseudo-code for encrypting each object using the encryption key for each object and generating a digital signature for the background image and the encrypted object is as shown in FIG. 7.

8 is a flowchart illustrating an image processing method according to an embodiment of the present invention. Each of the steps performed in the following is described with reference to FIGS. 1 to 7, each object is encrypted based on the encryption key for each object by the control device 110, and for the encrypted object and the background image A method of decoding and outputting only a specific encrypted object after the digital signature is generated will be described. Each step performed below may be performed in a separate image processing device 120 different from the control device 110.

In operation 810, the image processing apparatus 120 receives the encrypted frame and digital signature from the control apparatus 110. Here, the encrypted frame may be a frame in which only some objects are encrypted.

That is, as shown in FIG. 2E, the region of the object extracted in the image (frame) is masked. In addition, each extracted object is already encrypted by a separate encryption key, as described above.

In step 815, the image processing apparatus 120 verifies the digital signature of the encrypted frame using the public key.

If the verification fails, the process proceeds to step 810. If the verification fails, the image processing device 120 may output a guide message according to the verification failure, and then terminate the process without proceeding.

However, if the verification is successful, in step 820, the image processing apparatus 120 uses the decryption key to decrypt the object to be decoded.

In operation 825, the image processing apparatus 120 combines and outputs the decoded object and the background image. As already described above, the background image includes only information on the remaining areas except for the extracted object area. The object region extracted from the background image is already shown as being masked as described above. Accordingly, the image processing apparatus 120 allows only the decoded object to be viewed by being combined with the background image, and the remaining extracted encrypted object areas may be masked and output.

9 is a diagram schematically showing an internal configuration of an image processing apparatus according to an embodiment of the present invention, and FIG. 10 selectively decodes only a specific object according to an embodiment of the present invention and outputs it by combining it with a background image It is a diagram illustrating a pseudo code.

Referring to FIG. 9, the image processing apparatus 120 according to an embodiment of the present invention includes a key generator 910, a decoder 915, an output unit 920, a memory 925, and a processor 930. It includes.

The key generation unit 910 generates a decryption key for the object to be decrypted. This may be extracted based on the identification information (id) of the object to be decoded from the key set (K) for each object extracted from the image. In FIG. 9, the image processing apparatus 120 is shown to generate a decryption key, but the decryption key may be input by a key manager from the outside.

The decoding unit 915 decodes the object to be decoded using the decoding key.

Before the object to be decrypted is decrypted by the decryption unit 915, the digitally signed encrypted frame C may be verified based on the public key. It is natural that the verification process may be performed by the image processing device 120 or may be performed by a separate verification device. Here, the public key may be a part of a key pair (secret key, public key) generated for digital signature at a specific time point described in FIGS. 1 to 7.

The output unit 920 combines the decoded object with the background image and outputs it under the control of the processor 930.

The memory 925 selectively decodes only a specific object according to an embodiment of the present invention and outputs an image, thereby various algorithms, instructions, and processes for performing an image processing method that can protect the privacy of others Stores various derived data.

The processor 930 includes internal components (eg, a key generation unit 910, a decoding unit 915, an output unit 920, a memory () of the image processing apparatus 120 according to an embodiment of the present invention. 925) and the like.

The pseudo code for selectively decoding only a specific object and combining it with the background image to output it is as illustrated in FIG. 10.

The image authentication method according to the present invention described above can be implemented as computer readable codes on a computer readable recording medium. The computer-readable recording medium includes any kind of recording medium storing data that can be read by a computer system. For example, there may be a read only memory (ROM), a random access memory (RAM), a magnetic tape, a magnetic disk, a flash memory, and an optical data storage device. In addition, the computer-readable recording medium may be distributed over computer systems connected by a computer communication network, and stored and executed as code readable in a distributed manner.

The above-described embodiments of the present invention have been disclosed for purposes of illustration, and those skilled in the art having various knowledge of the present invention will be able to make various modifications, changes, and additions within the spirit and scope of the present invention. It should be regarded as belonging to the following claims.

110: control device
115: image processing device
200: video authentication device

Claims (12)

In the video authentication method in the video authentication device,
(a) extracting each object from the image;
(b) encrypting each extracted object with a different key; And
(c) generating a digital signature for each of the encrypted objects and the background image using a secret key at a specific time point,
Step (b) is,
Extracting different encryption keys for each object from the master key; And
And encrypting each object with a different key using the extracted encryption key for each object.
delete According to claim 1,
The background image is an image authentication method, characterized in that it is a region other than each of the extracted objects from the image.
According to claim 1,
And after the digital signature is generated, updating the secret key for digital signature at another time.
According to claim 1,
The digital signature is generated by using a forward secure signature algorithm.
In the video authentication device,
An object extraction unit for extracting each object from the image;
An encryption unit that encrypts each extracted object with a different key; And
It includes a signature generating unit for generating a digital signature for each of the encrypted object and the background image using a secret key at a specific time,
The encryption unit,
A video authentication device characterized by extracting different encryption keys for each object from a master key, and encrypting each object with different keys using different encryption keys extracted for each object.

The method of claim 6,
Further comprising a key generator for generating keys for object encryption and digital signature,
The key generation unit is a video authentication device, characterized in that for generating a master key used for each object encryption and a secret key and a public key pair for digital signature at a specific point in time.
delete The method of claim 7,
And an update unit for updating the secret key for digital signature at another time after generating the digital signature.


delete delete delete

Images