KR20200088236A - Trackable de-identification method, apparatus and system - Google Patents

Abstract

According to an embodiment of the present invention, a traceable de-identification method comprises the steps of: receiving original data including object identification information; generating de-identification data by de-identifying the object identification information from the original data; storing the original data and the de-identification data in a database; and tracking an object based on the de-identification data.

Description

TRACKABLE DE-IDENTIFICATION METHOD, APPARATUS AND SYSTEM}

The present invention relates to a trackable de-identification method, apparatus, and system, and more particularly, to a method, apparatus, and system capable of simultaneously de-identifying and tracking object identification information.

The de-identification technology of the data collected through the Internet of Things (IoT) device is to process the de-identification of the original data collected to prevent the leakage of personal information of data collected and shared through a number of things sensors. Technology. This de-identification technology has the advantage of preventing privacy infringement by removing elements that can identify individuals from original data such as video and audio, and ensuring confidentiality.

However, some of the existing de-identification technology has weak security because it is possible to restore the original, and the non-identification technology, which is not possible to restore the original, is anonymized all the information related to the object, so no information remains and big data is not left. There is a limit to the availability of furnaces.

In order to solve the above problems, the present invention is to provide a traceable de-identification method, apparatus, and system that simultaneously improves security as well as data utilization by applying de-identification technology and object tracking technology to original data simultaneously. do.

A traceable de-identification method according to an embodiment includes receiving original data including object identification information; Generating non-identification data by de-identifying the object identification information from the original data; Storing the original data and the de-identification data in a database; And tracking the object based on the de-identification data.

The generating of the de-identification data may include: recognizing the object to be de-identified; Extracting feature information of an object from the object identification information; And generating the de-identification data by allocating an identifier (ID) to the object.

The database includes a first database and a second database, and storing the original data and the de-identification data in a database comprises: storing characteristic information of the object and the identifier in the first database; And encrypting and processing the original data in the second database.

The original data includes image data, and the step of generating the de-identification data includes: distinguishing a de-identification region including a fixed region from the image data and a region in which the object moves; Detecting pixels of the moving object with respect to the de-identification area; And generating the de-identification data by allocating the identifier to the detected pixel.

The original data includes voice data, and the generating of the de-identification data comprises: selecting a de-identification area including the object identification information based on the context of the voice data; Removing the de-identification area from the voice data; And inserting a sound different from the existing sound of the voice data into the area where the non-identification area is removed.

The tracking of the object may include analyzing movement information of the object using the de-identification data, the feature information, and the identifier.

In the step of receiving the original data, the original data is received from a plurality of Internet of Things devices including a first Internet of Things device, and the trackable de-identification method comprises: a user using the user device as the first object Requesting information necessary for the Internet device; Searching for a peripheral device including an IoT device excluding the first IoT device according to the user's request; Receiving the de-identification data from the peripheral device; And responding to the user device based on the de-identification data.

A traceable de-identification apparatus according to an embodiment includes: a communication unit receiving original data including object identification information; A control unit operable to track an object using de-identification data generated from the original data; And a memory including a database in which the original data and the de-identification data are stored; and wherein the controller de-identifies the object identification information from the original data to generate the de-identification data. Angry engine; And a tracking engine that tracks the object based on the de-identification data.

The de-identification engine includes: an object recognition unit that recognizes the object to be de-identified; A feature detector for extracting feature information of the object from the object identification information; A generator configured to generate the de-identification data by allocating an identifier (ID) to the object; And a processing unit that stores the original data and the de-identification data in the database.

The database includes a first database and a second database, and the processing unit may store characteristic information of the object and the identifier in the first database, and encrypt and store the original data in the second database. .

The original data includes image data, and the de-identification engine includes a first data portion that de-identifies the image data to generate first de-identification data, and the first data portion is fixed from the image data. The area and the de-identification area including the area in which the object is moved are distinguished, the pixel of the moving object is detected with respect to the de-identification area, and the identifier is assigned to the detected pixel, thereby generating the De-identification data can be generated.

The original data includes voice data, and the de-identification engine includes a second data unit that de-identifies the voice data to generate second de-identification data, and the second data unit includes the voice data. A de-identification area including the object identification information is selected based on context, the de-identification area is removed from the speech data, and an area different from the existing sound of the speech data is removed from the de-identification area. The second de-identification data may be generated by inserting a sound.

The tracking engine may analyze movement information of the object using the de-identification data, the feature information, and the identifier.

A traceable de-identification system according to an embodiment includes a traceable de-identification device according to the above-described contents; An IoT device transmitting original data including object identification information to the de-identification device; And a user device requesting information required for the IoT device and receiving tracking data including movement information of the object from the de-identification device.

According to embodiments of the present invention, a traceable de-identification method, apparatus, and system that improves not only security but also data utilization possibility by simultaneously applying de-identification technology and object tracking technology to original data may be provided. .

1 is a diagram schematically showing a traceable de-identification system according to an embodiment of the present invention.
2 is a view more specifically showing a de-identification engine according to an embodiment of the present invention.
3 is a diagram more specifically showing a first database according to an embodiment of the present invention.
4 is a view for explaining an internal network communication method of a traceable de-identification system according to an embodiment of the present invention.
5 is a view for explaining an external network communication method of a traceable de-identification system according to an embodiment of the present invention.
6 is a flowchart illustrating a traceable de-identification method according to an embodiment of the present invention.
7 is a flow chart for explaining in more detail some steps of the traceable de-identification method according to FIG. 6.
8 is a flowchart for explaining a traceable de-identification method according to an embodiment of the present invention from a system perspective.

The present invention can be applied to various transformations and can have various embodiments, and specific embodiments will be illustrated in the drawings and described in detail in the detailed description. Effects and features of the present invention and methods for achieving them will be clarified with reference to embodiments described below in detail together with the drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various forms.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings, and the same or corresponding components will be given the same reference numerals when describing with reference to the drawings, and redundant description thereof will be omitted. .

In the following examples, terms such as first and second are not used in a limiting sense, but for the purpose of distinguishing one component from other components. In the following embodiments, the singular expression includes a plural expression unless the context clearly indicates otherwise. In the examples below, terms such as include or have are meant to mean that features or components described in the specification exist, and do not preclude the possibility of adding one or more other features or components. In the drawings, the size of components may be exaggerated or reduced for convenience of description. For example, since the size and shape of each component shown in the drawings are arbitrarily shown for convenience of description, the present invention is not necessarily limited to what is illustrated.

Each of the features of the various embodiments of the present invention may be partially or totally combined or combined with each other, and technically various interlocking and driving may be possible as those skilled in the art can fully understand, and each embodiment may be independently implemented with respect to each other. It can also be implemented together as an association.

1 is a diagram schematically showing a traceable de-identification system according to an embodiment of the present invention.

Referring to FIG. 1, a trackable de-identification system according to an embodiment includes an Internet of Things (IoT) device 100, a communication network 20, and a trackable de-identification device 10 (hereinafter, non- Identification device 10) and a user device 600.

The trackable de-identification system according to an embodiment may be an object de-identification and tracking solution targeting original data including video data, audio data, and the like. 1 is not limited to the number of the IoT device 100, the de-identification device 10, or the user device 600 as an example for explaining the present invention.

The IoT device 100 is a device that collects original data about an external environment, and may include a device that uses a camera and a microphone. The IoT device 100 may be connected to the de-identification device 10.

The original data may include object identification information capable of identifying objects to be de-identified and tracked. The object identification information may include personal information and sensitive information capable of identifying an individual, and may include various information including elements for identifying an individual.

For example, the IoT device 100 that collects video data may include a CCTV, a black box, etc., and the IoT device 100 that collects voice data may include an AI speaker, a recorder, and the like. The IoT device 100 is not limited to the above, and may include a device with a camera or a microphone that collects, stores, processes, and transmits one or more data of both video and audio data.

The IoT device 100 may be designed not only to be a single device, but also to connect and communicate with multiple devices.

The IoT device 100 may include a sensor 102. The sensor 102 is not limited to a camera or microphone, and further includes at least one of an infrared sensor, a temperature/humidity sensor, and a sound detection sensor to improve the quality of the collected data. The sensor 102 may be provided in the form of being embedded or attached to the IoT device 100, and the provided form of the sensor 102 does not limit the present invention.

The collected original data may directly use data stored in a separate database or user-specified data without going through the IoT device 100.

The original data includes, but is not limited to, image and/or audio data collected from a sensor 102-based camera and/or a device with a microphone attached thereto, Internet of Things (IoT) data, data collected from various peripheral sensors After de-identification of an object, various data necessary when used in a service linked with an object tracking service may be included. In addition, the original data may include various types of data, such as text and audio-visual recognition.

The traceable de-identification device 10 according to an embodiment may include a communication unit 11, a control unit 12, and a memory 13.

The communication unit 11 may receive original data including object identification information.

The communication unit 11 provides a function for the IoT device 100, the de-identification device 10, and the user device 600 to communicate with each other through the communication network 20, or other traceable de-identification systems (not shown) City).

The single or multiple IoT devices 100 use a wireless or wired communication method to communicate with each other and/or de-identify the device 100 or the user device 600 through the communication network 20. ).

For example, the communication network 20 includes a Personal Area Network (PAN), a Local Area Network (LAN), a Campus Area Network (CAN), a Metropolitan Area Network (MAN), a Wide Area Network (WAN), and a Broad Band Network (BBN). , It may include any one or more of the network, such as the Internet, but is not limited thereto.

The control unit 12 may operate to track an object using de-identification data generated from original data. The control unit 12 includes a de-identification engine 200 for de-identifying objects and a tracking engine 300 for tracking objects.

The controller 12 may include all types of devices capable of processing data, such as a processor. Here, a'processor' may mean a data processing device embedded in hardware having physically structured circuits, for example, to perform functions represented by codes or instructions included in a program. As an example of such a data processing device embedded in hardware, a microprocessor, a central processing unit (CPU), a processor core, a multiprocessor, and an application-specific integrated (ASIC) Circuit), FPGA (Field Programmable Gate Array), and the like, but the scope of the present invention is not limited thereto.

The de-identification engine 200 according to an embodiment may generate de-identified data (hereinafter, referred to as de-identification data) by de-identifying object identification information from the original data.

The de-identification engine 200 may remove object identification information from the original data by making it irretrievable. The de-identification engine 200 may assign a digital identifier (ID) to the de-identification data based on the characteristics of the object.

For example, when de-identifying image data, pixels of a moving object having a change in contrast to fixed data as a background may be detected and replaced with a non-identification symbol. When the voice data is de-identified, object identification information is determined based on the context of the voice, and then the corresponding portion can be basically deleted and replaced with a de-identified sound.

The detailed operation of the de-identification engine 200 will be described in more detail in FIG. 2 to be described later.

The tracking engine 300 according to an embodiment may track an object by analyzing movement information of the object based on de-identification data, an identifier, and the like.

The tracking engine 300 may recognize and discriminate multiple moving objects in an image and continuously track and analyze moving information and behavior of the object based on the digital identifier assigned to each object. To collect object movement pattern information, zoom-in, zoom-out, screen switching, and direction change can be used. The tracking engine 300 may continuously track the object determined to be the same moving object by using the video and audio collected from the plurality of IoT devices 100 together.

When tracking an object using the tracking engine 300, the sensor 102 may be a single sensor, or multiple sensors may operate in cooperation as multiple sensors.

The memory 13 functions to temporarily or permanently store data processed by the de-identification device 10.

The memory 13 may include a database (hereinafter referred to as a DB) in which original data, de-identification data, and/or data in which these data are processed are stored. The memory 13 may include a first database 400 storing de-identification data and a second database 500 storing original data. The first database 400 may be referred to as an'open database', and the second database 500 may be referred to as an'encryption database'.

The de-identification data may be stored in the open database 400 and disclosed to the outside. The digital identifier (ID) of the de-identified object and characteristic information of the object may be stored in the open database 400. The number of the feature information is not limited, and can be increased infinitely as needed. And, in the open database 400, grouping information of objects grouped with similar features may be stored. Data stored in the open database 400 may be continuously updated by the de-identification engine 200 and the tracking engine 300.

The open database 400 will be described in more detail in FIG. 3 to be described later.

The encryption database 500 may store original data collected from the outside (eg, the IoT device 100). The original data may be encrypted and stored safely through encryption processing, and manually added data may be further included. The encryption processing method may be stored by being bidirectionally encrypted or unidirectionally encrypted depending on the type and sensitivity of data.

Depending on the embodiment, the digital identifier and feature information assigned to the object may be stored separately in separate databases from the first and second DBs 400 and 500. In addition, the memory 13 according to an embodiment may include both the open database 400 and the encryption database 500, or may include only one of the two.

The memory 13 may include a magnetic storage media or a flash storage media, but the scope of the present invention is not limited thereto.

The de-identification device 10 de-identifies the data collected by the IoT device 100 at the request of the user device 600 using the de-identification engine 200 and de-identifies the data. Objects can be tracked by utilizing the tracking engine 300. Here, the utilization data may include at least some of de-identified image data, audio data, feature information, grouping information, etc. stored in the open database 400, but is not limited thereto, and the original data stored in the encryption database 500 May be included.

As described above, according to an embodiment of the present invention, it is possible to improve not only security but also data utilization possibility by simultaneously applying de-identification technology and object tracking technology to the original data.

2 is a view more specifically showing a de-identification engine according to an embodiment of the present invention.

The de-identification engine 200 according to an embodiment may include a first data unit 210 and a second data unit 220. The data units 210 and 220 may include object recognition units 212 and 222, feature detection units 214 and 224, generation units 216 and 226, and processing units 218 and 228, respectively.

The object recognition units 212 and 222 may recognize an object to be de-identified. The feature detection units 214 and 224 may extract feature information of the object from the object identification information. The generating units 216 and 226 may generate the de-identification data by assigning an identifier (ID) to the corresponding object. The processing units 218 and 228 may store and manage original data, de-identification data, and feature information for each detected object in a database.

Hereinafter, it will be described as an example that the first data unit 210 is an image data unit in charge of image data, and the second data unit 220 is an audio data unit in charge of audio data.

The first data unit 210 may recognize the object based on the feature, extract the feature of the object, and generate the first de-identification data and store it in the first database 400.

Specifically, the first object recognition unit 212 may receive the collected image data and recognize an object that is a non-identification target in the image. Objects are not limited to humans, and may include objects that may be infringed on personal information. Objects subject to personal information infringement may include a vehicle number, a personal mobile phone number, and sensitive information.

The first feature detection unit 214 may detect feature information of the recognized object and select it as a non-identification target. The number of the feature information is not limited, and can be detected as necessary for object identification. Feature information may be temporarily stored in a memory (not shown) mounted on the de-identification engine 200.

The first de-identification unit 216 may designate various types of information that can be used to identify a moving object or an individual that is a non-identification target in an image as a non-identification target. The object (personal) identification information may include an object's body information, personal information, and sensitive information.

The first non-identification unit 216 may distinguish a fixed area and a non-identification area from image data. The de-identification area may include an area in which an object to be de-identified moves. Thereafter, the first de-identification unit 216 may detect a pixel corresponding to the moving object with respect to the de-identification area. The first de-identification data may be generated by assigning a digital identifier to the detected pixel. The detected pixel may be removed and the removed pixel may be replaced to insert and replace the de-identification symbol.

The de-identified object is assigned a digital identifier so that it can be used to track the object later.

The first processing unit 218 may store the first de-identification data, the feature information detected by the first feature detection unit 214, and the digital identifier in the open database 400, or may be processed and stored. At this time, the original data may be encrypted and stored in the encryption database 500. When storing information, in order to efficiently track objects, objects having the same characteristics may be grouped and stored using grouping information, but in addition, numerical values obtained through a specific calculation formula may be additionally stored.

The second data unit 220 may recognize an object based on context, extract features, and generate second de-identification data and store it in the first database 400.

Specifically, the second object recognition unit 222 may receive the collected voice data and recognize an object that is an object of non-identification in voice.

The second feature detection unit 224 may detect a feature of a portion corresponding to personal information in a voice and select it as a non-identification target. The number of feature information detected by the second feature detector 224 is not limited and can be detected as necessary for object identification. The identified features may be temporarily stored in a memory (not shown) mounted on the de-identification engine 200.

The collected voice may designate various information that can be used to identify an individual in the voice in the second de-identification unit 226 as a de-identification target. The object identification information may include the object's phone number, address, and sensitive information.

The second de-identification unit 226 may select a de-identification area including object identification information based on the context of the voice data, and remove the deactivation area from the voice data. The second de-identification unit 226 may generate second de-identification data by inserting a sound different from the existing sound of the speech data in the area where the inactive area is removed.

The second processing unit 228 may store each feature in the open database 400. The second processing unit 228 may process and store the second de-identification data, the feature information detected by the second feature detection unit 224, and the like in the open database 400. At this time, the original data may be encrypted and stored in the encryption database 500. Also, the second processing unit 228 may group and store objects having the same characteristics using grouping information.

The specific de-identification method of the data units 210 and 220 is not limited to the above, and may vary depending on the type of the collected data.

3 is a diagram more specifically showing the first database 400 according to an embodiment of the present invention.

In the open database 400, de-identification data 410, feature tables 420, and grouping information 430 may be stored.

The de-identification data 410 may include data in which object identification information is de-identified from the original data.

In the feature table 420, the digital identifier (ID) 422 and the feature information 424 of each object detected by the de-identification engine 200 may be stored, but is not limited thereto. Information about frequency, location, etc. may also be stored. The data 422 and 424 stored in the feature table 420 may be used to effectively track the object in the tracking engine 300 later.

The feature table 420 may be continuously updated by the de-identification engine 200 and the tracking engine 300 to keep up to date. Data that has not been used for a long period of time in the feature table 420 may be periodically deleted to increase accuracy and availability for an object appearing in an image.

The feature information 424 stored in the feature table 420 is classified into short-term, mid-term, and long-term data and differentially reflects features required to identify an object in consideration of the environment in which the traceable de-identification device 10 is used. Can. Short-term features may include clothes, shoes, ornaments, and accessories worn by the object. Mid-term features may include object movement patterns, appearance times, frequency of appearance, and the like. Long-term features may include physical features, such as the face of the object, and age.

If there is an object that appears fixedly in the image when analyzing the object movement pattern, appearance time, and frequency of occurrence of the medium information among the feature information 424, a cache memory (not shown) located outside the open database 400 is used. It can quickly identify objects and reduce the false positive rate.

The grouping information 430 is information that is referenced so that the tracking engine 300 can efficiently track objects. When tracking the designated object in the tracking engine 300, and it is difficult to track only the digital identifier (ID), the grouping and coloring function using the grouping information 430 may be utilized. Grouping is a function to manage and manage objects or objects to be tracked as a group, and coloring can be a function to increase visibility in object tracking by designating a tracking object as a desired color.

For example, if you need to monitor and track target groups that you want to control in prisons, control zones, etc., or if you need to safely protect target groups in daycare centers or kindergartens, sample targets to be tracked among multiple objects and group them in special colors. Can be displayed. In this case, when an abnormal behavior is detected from a control target, it may be displayed in a different color or sounded by a warning sound.

As another example, when tracking consumption patterns and travel routes of a target group in a commercial area, an offline store, an exhibition, etc., each color is displayed according to a desired criterion such as an object's gender and age, and a filter option function is provided to classify each interest group. You can add

4 is a view for explaining an internal network communication method of a traceable de-identification system according to an embodiment of the present invention.

The traceable de-identification device 10 connected by one network may be configured by connecting a plurality of IoT devices 100 to the same communication network 20. In other words, the plurality of IoT devices 100 and the de-identification device 10 are connected by the same communication network 20 and can exchange information with each other through the communication network 20.

Each IoT device 100 may communicate with each other through a sensor 102, exchanging and receiving original data generated in a space in which each device is in charge. When original data is collected and used from a plurality of devices 100, it is possible to accurately determine whether objects appearing in video and audio are the same object.

In addition, a plurality of IoT devices 100 can cooperate to collect, process, and store data to efficiently process data using the performance, time, and battery of the devices 100. By arranging the plurality of IoT devices 100, it is possible to reduce the error range of the original data and accurately grasp the movement line of the moving object.

According to an embodiment, the user may selectively use the de-identification engine 200 and the tracking engine 300 in the traceable de-identification device 10. For example, when only the tracking engine 300 is to be used, the fingerprint recognition sensor 102 installed in the same communication network 20 confirms the entry and exit of the object and de-identifies the device 10 through the IoT device 100. ). In this case, the de-identification engine 200 may be partially used only when entering, and the de-identification engine 200 may not be used when leaving. Alternatively, the presence or absence of an object may be determined using the infrared sensor 102, and when the object to be tracked is absent, the operation of the devices 100 is stopped, thereby unnecessary in an environment in which a plurality of IoT devices 100 are installed. Energy efficiency can be increased without collecting images.

As another example, when the IoT device 100 including the camera includes the motion recognition sensor 102, even if sufficient data to identify the object does not accumulate in the databases 400 and 500, they are in the same communication network 20. The devices 100 may cooperate to detect a user's motion and predict an object's movement pattern and motion pattern.

5 is a view for explaining an external network communication method of a traceable de-identification system according to an embodiment of the present invention.

The traceable de-identification device 10 may operate in conjunction with an external IoT device 110 located outside the communication network 20. When a user requests specific data obtained through communication with the outside to the IoT device 100 through the user device 600, the sensor 102 of the device 100 is an external sensor of the external IoT device 110 By sending a signal to (112) it can receive the data required for the user request.

The IoT device 100 controlled by the user can be built on the same communication network 20 as the trackable de-identification device 10, and the sensor 102 attached to the device 100 is an external sensor 112. It can be designed in a structure to communicate with.

Depending on the embodiment, the traceable de-identification device 10 may directly request and communicate with the sensor 112 of the external IoT device 110.

6 is a flowchart illustrating a traceable de-identification method according to an embodiment of the present invention. The traceable de-identification method according to an embodiment of the present invention described below may be performed in the de-identification apparatus 10 described above. Hereinafter, the contents overlapping with the above-described contents in FIGS. 1 to 5 will be briefly or omitted, and will be described with reference to FIGS. 1 to 5 together.

Referring to FIG. 6, original data including object identification information may be received (S100 ). The IoT device 100 equipped with the sensor 102 may collect original data including various data such as voice and video.

Thereafter, de-identification data may be generated by de-identifying object identification information from the original data (S200 ). The de-identification engine 200 may de-identify object identification information from the original data in a manner suitable for each data type.

When the original data is image data, step S200 may include the following steps.

A non-identification area including a fixed area and an area in which the object moves may be distinguished from the image data. Subsequently, a pixel of an object moving with respect to the de-identification area may be detected, and an identifier 422 may be assigned to the detected pixel to generate de-identification data 410. At this time, an identifier 422 may be allocated to the detected pixel, and replaced with a non-identifying symbol.

Meanwhile, when the original data is voice data, step S200 may include the following steps.

Based on the context of the speech data, a de-identification region including the object identification information may be selected, and the de-identification region may be removed from the speech data. Thereafter, a sound different from the existing sound of the voice data may be inserted into the area where the non-identification area is removed to replace it.

Thereafter, the original data and the non-identifying data may be stored in the databases 400 and 500 (S300). The original data and de-identification data may be stored in the databases 400 and 500. At this time, the original data may be stored in the encryption database 500 through encryption processing.

Thereafter, an object may be tracked based on the de-identification data (S400). The tracking engine 300 may track the object based on the feature information 424 of the object extracted in step S200. Object tracking may be performed with only a single IoT device 100 or may be performed using a plurality of IoT devices 100.

Referring to FIG. 4 again, a traceable de-identification method according to an embodiment of the present invention will be described. In step S100 of receiving original data, the original data may be received from a plurality of IoT devices 100 including a first IoT device 100a. At this time, the step of generating de-identification data (S200) may include the following steps.

The user may request information required for the first IoT device 100a using the user device 600. At the request of the user, a peripheral device including the IoT device 100 except for the first IoT device 100a may be searched. The peripheral device may include the IoT device 100 excluding the first IoT device 100a and the external IoT device 110 (see FIG. 5 ). Thereafter, de-identification data may be received from the peripheral device, and a user device may be responded to based on the de-identification data. The response may be to receive object tracking information generated based on de-identification data.

7 is a flowchart for explaining in more detail the step S200, which is a part of the traceable de-identification method according to FIG. 6. The step of generating de-identification data (S200) may include steps described below. It may be described with reference to FIGS. 1 to 3 together.

The object to be de-identified may be recognized (S210), and feature information 424 of the object may be extracted from the object identification information (S220). Then, the identifier (ID) may be assigned to the corresponding object to generate de-identification data 410 (S230).

The de-identification data 410 and the feature information 424 of the object extracted from the original data may be stored in the feature table 420 of the first database 400. At this time, the feature table 420 also stores digital identifiers (IDs) and grouping information 430 assigned to each object based on the feature information 424 of the object, so that the object can be collectively managed.

The step of tracking the object (S400) (see FIG. 6) may include analyzing movement information of the object using the de-identification data 410, feature information 424, and an identifier.

8 is a flowchart for explaining a traceable de-identification method according to an embodiment of the present invention from a system perspective. Hereinafter, with reference to FIGS. 4 and 5 together, a method of obtaining desired data using a de-identification device 10 that can be tracked according to a user's request will be described.

Referring to FIG. 8, a user may request required information through the first IoT device 100a (see FIG. 4) using the user device 600 (S600 ). Here, the IoT device 100 that recognizes the user's request may include a sensor 102 that recognizes a request such as voice recognition, image recognition, and touch recognition. For example, various IoT devices 100 such as home appliances, mobile equipment, and wearable devices may be connected to the Internet to recognize a user's request.

The IoT device 100 may search for a suitable peripheral device according to a user's request (S700). The peripheral device may be a device in which the trackable de-identification device 10 is interlocked, or may be a general Internet of Things device not interlocked with the de-identification device 10.

The IoT device 100 may receive de-identification data from a peripheral device having data requested by the user (S800).

The IoT device 100 receiving the data requested by the user may respond to the user device 600 based on the de-identification data (S900 ). At this time, the traceable de-identification device 10 can analyze and learn the command and response history based on the accumulated data by repeating the user's request and the response of the device 100, and providing an improved user-customized service can do.

Example

Hereinafter, exemplary embodiments that can be used in various fields such as safety, communication, transportation, health, medical, commercial, and insurance will be described.

The present invention can provide a system for a user who needs indoor monitoring, such as furniture for raising companion animals, double-edged furniture, single-person furniture, and the like to safely manage the situation inside the house. Referring to FIG. 4, the IoT device 100 equipped with the sensor 102 may be a home CCTV in the home. By installing an Internet of Things (100) device such as a home CCTV, a home camera equipped with a camera, an IP camera, etc. in the home to control the direction of the camera to receive the situation inside the house in real time and store it in the cloud storage through the communication network (20) Can be used in any way.

At this time, the possibility of leakage of personal information and sensitive information that may occur due to data collected in a house where the personal privacy is most exposed is solved through the de-identification engine 200 of the de-identification device 10 of the present invention. Can. Since the image de-identification unit 216 of the de-identification engine 200 replaces pixels considered as privacy among the collected image data with non-identification symbols, it is impossible to restore the original data. IP camera) Even if data is leaked to the outside, it can be safe for security. In addition, the voice de-identification, which removes the part from the voice de-identification unit 226 and replaces it with other sounds, proceeds with the voice de-identification unit 226 for voice information that can cause a serious blow to an individual if it leaks, such as sensitive information or account information collected through the microphone Can.

After identifying a family member based on the characteristics of an object collected through a camera and a microphone, which can be the sensor 102 of the IoT device 100, and then giving a unique digital ID, the moving object itself in the original data cannot be de-identified. Can. Here, if the detailed characteristics between members are recognized and recognized as different people, they may be processed as different symbols. The technology for classifying personal information and sensitive information will be more effectively learned through training data, and the performance of the de-identification engine 200 and the tracking engine 300 may be continuously improved through test data.

When a plurality of cameras 100 are installed in a house as shown in FIG. 4, the tracking engine 300 can accurately track moving objects through networking between each camera 100. In an environment with a single camera 100, an object may be tracked using a zoom-in or zoom-out function. Accurate analysis and learning are possible as the data of the object accumulates, and customized services optimized for the user can be provided.

In addition, the present invention can provide an appropriate advice to the user about the electricity or gas that is wasted unnecessarily, by grasping the usage status of the home appliance through sensors mounted in the smart home. The sensor 102 attached to various smart home appliances 100 may be an image sensor, a voice sensor, an infrared sensor, a temperature/humidity sensor, a sound detection sensor, and the like. Through the sensor 102, it is possible to grasp the current state of electricity and gas use in the home and provide a service that can prevent a wasted portion.

In addition, the present invention can be applied to an in-vehicle audio processing device, a roadside, a roadside, and a CCTV installed on a parking lot to provide users with indoor traffic information such as an indoor parking lot that is difficult to grasp with existing satellite information. Referring to FIG. 5, when a user requests a query such as'tell me about the current traffic condition' or'tell me about the free space in the parking lot' through the voice assistant 102 mounted in the vehicle in the car, the voice assistant immediately prompts the user Search the parking lot in the requested area to inform you of the situation. At this time, the in-vehicle IoT device 100 may recognize the corresponding voice and communicate with the external IoT device located in the vicinity of the CCTV 110 to obtain information. The external IoT device 110 may be a device having traffic information or other information, such as CCTV installed indoors, in addition to CCTV installed on roads and roadsides. Through networking with the peripheral device 110, the in-vehicle IoT device 100 may receive indoor traffic and more accurate traffic information that are difficult to grasp with existing satellite information. If there is a parking lot preferred by the user, the parking situation may be transmitted by communicating with the CCTV 110 or the external sensor 112 in the corresponding parking lot. At this time, the voice command stored in the vehicle in-memory may be safely de-identified through the de-identification engine 200 of the de-identification device 10. Thereafter, in order to improve the quality, the voice de-identification unit 226 may remove the information capable of identifying the individual from the user query, and then de-identify and store it in the open database 400. Thereafter, when a new request is received from the user, a customized service can be provided by using the existing information stored in the open database 400 together.

In addition, object identification information among video and audio data collected by the black box 100 installed in the vehicle may be de-identified. These technologies can also be used in video and audio processing devices used in autonomous vehicles and other industrial institutions in the future.

In addition, the present invention can track the movement of the elderly through the IoT device 100 of a home or nursing home where the elderly live alone and analyze the movement paths and patterns to prepare for emergencies and emergencies. Referring to FIG. 4, when a plurality of IoT devices 100 such as a camera are mounted at each location where the elderly frequently stay, it is possible to grasp the path of the elderly moving at home. The tracking engine 300 of the de-identification device 10 applied to the installed multiple cameras 100 may collect data according to the movement path of the elderly through networking between devices. At this time, since it is possible to analyze the exact movement form and pattern of the elderly, it may be possible to provide appropriate care and customized services. The guardian who wants to observe the movement of the elderly can check the safety of the elderly through the user device 600. If an abnormality different from a characteristic stored in an open database 400 in which an elderly person's behavior or movement patterns are recorded is captured, for example, the user and the guardian may be provided with the specifics through networking between the device in the kitchen and the device in the living room. Can provide. In addition, the IoT device 100 may provide a more accurate service by collecting and sharing not only image data but also voice data. Since the image data collected by the IoT device 100 that identifies the behavior of the elderly is through the image de-identification unit 216 and the voice data is originally de-identified through the voice de-identification unit 226, the safety of the service is Can be secured.

In addition, the present invention can technically solve the privacy problem of operating room imaging by blocking it with an external system by de-identifying moving objects such as doctors, nurses, patients, etc. in CCTV image data in the operating room of the medical industry. By de-identifying the object in the image through the de-identification engine 200 and tracking the object with the tracking engine 300, the original data is confirmed by checking the internal situation through the IoT device 100 with a camera installed in the operating room. Can be actively used.

Since medical data stored in the open database 400 is de-identified and replaced with a symbol, it can be disclosed to the outside to reduce disputes during medical accidents, and to investigate the exact cause of accidents in the event of an accident. In addition, since it is possible to collect sensitive medical data that was not possible before, it is possible to provide customized medical services to guardians, medical staff, and caregivers in the future, and to provide improved services such as the development of treatment for incurable diseases.

As shown in FIG. 4, when a plurality of IoT devices 100 such as CCTV are connected to a hospital room, a treatment room, and an operating room to communicate with each other, accuracy of a patient and a medical staff can be improved. When grouping and coloring are applied to the present embodiment, objects of interest can be set to efficiently track objects. For example, the grouping information 430 (refer to FIG. 3) may be used to classify a group of medical staff and hospital administrators and patients, and designate and manage each group in a different color.

In addition, the present invention can be applied to commercial areas with a large population of liquidity, such as shopping malls, street shopping districts, shopping districts, and plazas, and analyze commercial movement patterns to build commercial data.

Install IoT devices 100 such as CCTV in commercial areas, analyze movement patterns according to age and gender, and stay time from data collected by devices to capture consumption patterns and target group characteristics and use them for marketing or promotion Can effectively grasp the market size. Particularly, when a plurality of CCTVs installed on the street are not installed and consent to use video and audio data is not received by everyone, object identification information of original data using the de-identification device 10 of the present invention By de-identifying the data, security can be secured and used as big data. As data is collected for a long period of time from the IoT device 100 and the data is accumulated, accuracy of the corresponding data is improved and reliability of information can be provided.

When image data is collected through the CCTV (100) installed in a commercial area, each object is identified based on physical characteristics such as face, gender, and height that can be identified in the object in the image, and a digital ID is given, and replaced with a symbol to identify it. Can be angry. When the target object is designated from the de-identified image data, it is activated, outputs the digital ID of the object, and compares the data of the object with data stored in the open database 400 of the system to calculate how much it matches. After that, you can print them together. The new data for the object is recorded in the database 400 again, and is reflected when the tracking engine 300 compares the data in the existing database.

In addition, the present invention can clearly determine the cause of the accident and the negligence through de-identified video and audio data without privacy infringement in the insurance field, which requires appropriate evidence or data to prove objective facts.

In addition, the present invention can also be used in the insurance field that requires appropriate evidence or objective evidence. When an insured person claims an insurance claim to an insurance company, it is necessary to submit evidence to prove the accident. At this time, the de-identification device 10 according to an embodiment of the video and audio data collected from the various IoT devices 100 Can be applied and submitted as proof of accident.

If the de-identification device 10 is installed in a place where accidents occur frequently and in areas where accidents are expected, it is possible to clearly determine the cause and negligence of the accident without infringement of privacy. Can be reduced.

The embodiment according to the present invention described above may be implemented in the form of a computer program that can be executed through various components on a computer, and such a computer program can be recorded on a computer-readable medium. At this time, the medium may be a computer executable program. Examples of the medium include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and magneto-optical media such as floptical disks, And program instructions including ROM, RAM, flash memory, and the like.

Meanwhile, the computer program may be specially designed and configured for the present invention or may be known and available to those skilled in the computer software field. Examples of computer programs may include not only machine language codes produced by a compiler, but also high-level language codes executable by a computer using an interpreter or the like.

In addition, although the preferred embodiments of the present invention have been shown and described above, the present invention is not limited to the specific embodiments described above, and the technical field to which the present invention pertains without departing from the gist of the present invention claimed in the claims. In addition, various modifications can be made by those skilled in the art, and these modifications should not be individually understood from the technical idea or prospect of the present invention.

Therefore, the spirit of the present invention should not be limited to the above-described embodiments, but not only the claims to be described later, but all ranges that are equivalent to or equivalently changed from the claims of the present invention. It would be said to fall into the category.

10: de-identification device 100: IoT device
102: sensor 110: external IoT device
112: external sensor 11: communication unit
12: control unit 13: memory
20: network 200: de-identification engine
212, 222: object recognition unit 214, 224: feature detection unit
216, 226: generating unit 218, 228: processing unit
300: tracking engine 400: first database
410: de-identification data 420: feature table
422: digital identifier 424: feature information
430: grouping information 500: second database
600: user device

Claims (14)

Receiving original data including object identification information;
Generating non-identification data by de-identifying the object identification information from the original data;
Storing the original data and the de-identification data in a database; And
And tracking the object based on the de-identification data. According to claim 1,
The step of generating the de-identification data,
Recognizing the object to be de-identified;
Extracting feature information of an object from the object identification information; And
And allocating an identifier (ID) to the object to generate the de-identification data. According to claim 2,
The database includes a first database and a second database,
The step of storing the original data and the de-identification data in a database,
Storing characteristic information of the object and the identifier in the first database; And
And encrypting the original data and storing the encrypted data in the second database. According to claim 2,
The original data includes image data,
The step of generating the de-identification data,
Distinguishing a de-identification area including a fixed area and an area in which the object moves from the image data;
Detecting pixels of the moving object with respect to the de-identification area; And
And generating the de-identification data by allocating the identifier to the detected pixel. The method of claim 2 or 4,
The original data includes voice data,
The step of generating the de-identification data,
Selecting a de-identification area including the object identification information based on the context of the voice data;
Removing the de-identification area from the voice data; And
And inserting a sound different from the existing sound of the speech data into the area where the de-identification area is removed. According to claim 2,
Tracking the object,
And analyzing movement information of the object using the de-identification data, the feature information, and the identifier. According to claim 2,
In the step of receiving the original data, the original data is received from a plurality of IoT devices including a first IoT device,
The traceable de-identification method,
A user requesting information necessary for the first IoT device using a user device;
Searching for a peripheral device including an IoT device excluding the first IoT device according to the user's request;
Receiving the de-identification data from the peripheral device; And
And responding to the user device based on the de-identification data. A communication unit receiving original data including object identification information;
A control unit operable to track an object using de-identification data generated from the original data; And
Includes a memory including a database in which the original data and the de-identification data are stored.
The control unit,
A de-identification engine that de-identifies the object identification information from the original data to generate the de-identification data; And
And a tracking engine that tracks the object based on the de-identification data. The method of claim 8,
The de-identification engine,
An object recognition unit recognizing the object to be de-identified;
A feature detector for extracting feature information of the object from the object identification information;
A generator configured to generate the de-identification data by allocating an identifier (ID) to the object; And
And a processing unit configured to store the original data and the de-identification data in the database. The method of claim 9,
The database includes a first database and a second database,
The processing unit,
The characteristic information of the object and the identifier are stored in the first database,
A traceable de-identification device that encrypts the original data and stores it in the second database. The method of claim 9,
The original data includes image data,
The de-identification engine includes a first data unit that de-identifies the image data to generate first de-identification data,
The first data portion,
A non-identifying area including a fixed area and an area in which the object moves is distinguished from the image data,
Detecting pixels of the moving object with respect to the de-identification area,
A traceable de-identification device that generates the first de-identification data by allocating the identifier to the detected pixel. The method of claim 9 or 11,
The original data includes voice data,
The de-identification engine includes a second data unit that de-identifies the voice data to generate second de-identification data,
The second data portion,
A non-identification area including the object identification information is selected based on the context of the voice data,
Remove the de-identification area from the voice data,
A traceable de-identification device that generates the second de-identification data by inserting a sound different from the existing sound of the speech data in the area where the de-identification area is removed. The method of claim 9,
The tracking engine,
A traceable de-identification device that analyzes the movement information of the object using the de-identification data, the feature information, and the identifier. A traceable de-identification device according to any one of claims 8 to 13;
An IoT device transmitting original data including object identification information to the de-identification device; And
And a user device that requests information necessary for the IoT device and receives tracking data including movement information of the object from the non-identifying device.

Images