KR20220032244A - Disinfection system using thermal image and visible image - Google Patents

Abstract

The present invention relates to a disinfection system using a thermal image and a visible image. A camera and a thermal imaging camera are used to double-detect a person having fever and a person not wearing a mask among visitors and simultaneously display warning information to the outside upon occurrence of event, thereby effectively preventing the transmission of infectious diseases such as viruses and bacteria from the outside to maximize heat dissipation efficiency. The image obtained by the camera is analyzed to detect the facial pattern characteristics of the visitor in which facial pattern features, which are pattern features of the upper region centered on the eyes of the face, are detected, and then user authentication is performed by comparing with previously registered registration DBs so that the authentication is possible without the person having to take off the mask separately to perform user authentication, thereby increasing the convenience and efficiency of the certification process, and effectively preventing unnecessary authentication time delays caused by wearing and separating masks. Components such as cameras, thermal imaging cameras, speakers, and LED lights are manufactured as an integrated unit to facilitate the miniaturization and reduce manufacturing and operating costs.

Description

Disinfection system using thermal image and visible image

The present invention relates to a quarantine system using thermal images and images, and more specifically, a camera and a thermal imaging camera detect whether an occupant has a fever and wear a mask, and at the same time perform user authentication by utilizing the facial pattern features of entrants When authentication fails, or when a person with a fever or a person not wearing a mask is detected, warning information is displayed to the outside, allowing double detection and enforcement of a person with a fever or not wearing a mask, effectively preventing the transmission of viruses and bacteria from outside It relates to a quarantine system using thermal images and images that can maximize the quarantine efficiency.

Recently, highly contagious infectious diseases such as Corona 19, Middle East Respiratory Syndrome (MERS), Severe Acute Respiratory Syndrome (SARS), and avian influenza (AI) have been diversified. As it appears, various studies on quarantine systems to prevent the spread of infectious diseases are being conducted.

These infectious diseases are expressed in the body in various ways, but since they are most often accompanied by high fever, thermal imaging cameras are installed at the entrances of airports, various buildings, government offices, hospitals, etc. There is an increasing trend in the introduction of quarantine systems that detect people with high fevers among those who enter through the system.

In general, since these infectious diseases are transmitted rapidly through human saliva, breathing, coughing, etc., wearing a mask is being considered as the most basic measure to significantly lower the contagiousness and transmission rate of infectious diseases.

However, the conventional quarantine system can detect a fever among entrants using a thermal imaging camera, but has a disadvantage in that the evacuation efficiency is lowered because the technology and method for detecting whether the entrants are wearing a mask are not configured at all.

On the other hand, with the recent rise in importance of security along with informatization, biometric technology and the development of electronic device industries, a user authentication system that allows only authorized users to access a specific space This is widely used.

In particular, since facial patterns have characteristics that are not only unique to each person, but also have no fluidity, research on a facial authentication system that authenticates users using facial patterns, which is a physiological characteristic unique to humans, is being conducted in various ways.

However, as the importance of wearing a mask has recently increased, most of the visitors wear a mask, and accordingly, when performing user authentication through the face authentication system, the visitor takes off the mask separately and separates it from the face to take a picture, When the authentication is completed, it is necessary to perform the operation of putting the separated mask back on the face, so that the authentication operation is cumbersome and complicated, the authentication time is delayed, and the convenience and efficiency of use are reduced.

In other words, 1) double detection and enforcement of a person who has a fever or not wearing a mask among the entrants, 2) authentication is possible without the user having to remove the mask separately during user authentication, and 3) quarantine that can maximize the user authentication speed There is an urgent need to study the system.

1 is a diagram showing the construction of a fever detection system disclosed in Korean Patent Registration No. 10-2091121 (Title of the Invention: Arrivals Fever Detection System).

The fever detection system (hereinafter referred to as the prior art) 100 of FIG. 1 includes a plurality of thermal cameras (C1 to C12) and a plurality of thermal cameras (C1-C12) for detecting the body temperature of an entrant moving along an entry route A central control unit 110 that collects, processes and monitors the images taken by C12), and a thermal camera test unit ( not shown).

At this time, the prior art 100 assumes a situation in which entrants move sporadically from a plurality of different gates (G1, G2~), and a plurality of thermal cameras (C1~C12) are used to detect entrants moving along the entry route. It has the form of continuous shooting.

The central control device includes an alarm that generates an alarm when the body temperature of the entrant detected by the plurality of thermal cameras C1 to C12 is in the preset reference temperature range.

The prior art 100 configured as described above collects data over the entire length of the entry flow of the entrant by installing a plurality of thermal cameras (C1 ~ C12) spaced apart along the entry route of the entrant to collect body temperature information for all entrants. It can be obtained easily, and it has the advantage of preventing the spread of endemic diseases abroad in Korea through the detection of fever of inbound travelers.

However, when the prior art 100 is applied to the entrance of a building, it is possible to detect a fever among the entrants using the thermal cameras C1 to C12, but the technology and method for detecting a person not wearing a mask are not described at all. Because of this, there is a problem that the prevention rate and quarantine efficiency of infectious diseases are lowered.

Even if the mask recognition system and the face authentication system are applied to the prior art 100, in the prior art 100, in order to perform mask recognition and face authentication, authentication must be performed in a state in which the mask of the entrant is detached. Not only is it cumbersome and complicated, but also the authentication time is delayed, and the convenience and efficiency of use are lowered.

In addition, since the prior art 100 simply compares the facial pattern features detected through facial imaging with previously registered registration information in a fixed order or random, respectively, the larger the number of registered registration information, the greater the There is a problem that the authentication time increases.

The present invention is to solve this problem, and the problem of the present invention is to use a camera and a thermal imaging camera to double detect a person with a fever and not wear a mask among the entrants, and at the same time to display warning information to the outside when an event occurs This is to provide a quarantine system using thermal images and images that can effectively prevent the transmission of infectious diseases such as viruses and bacteria from the outside and maximize the heat dissipation effect.

In addition, another solution to the present invention is to analyze the image acquired by the camera to detect the facial pattern features of the entrant, but after detecting the facial pattern features, which are the pattern features of the upper region centered on the eyes of the face, As it is configured to perform user authentication in comparison with the registered registration DBs, authentication is possible without requiring a separate mask to be removed to perform user authentication. This is to provide a quarantine system using thermal images and images that can effectively prevent delay in authentication time.

In addition, another solution to the present invention is that components such as a camera, a thermal imaging camera, a speaker, and an LED light are integrally manufactured, so miniaturization is possible as well as using thermal images and images that can reduce manufacturing and operating costs. This is to provide a quarantine system.

In addition, another solution of the present invention is to significantly increase the authentication speed and prevent unnecessary computational processing by configuring the registration DBs to be compared and compared in response to the user's access status during user authentication. This is to provide a quarantine system using thermal images and images.

A solution of the present invention for solving the above problems is a quarantine system including an access detector installed at an entrance of an indoor space to detect occupants: the entrance detector includes a bracket; and a device for photographing that is rotatably coupled to the bracket, wherein the device for photographing includes: a camera for photographing a person's face to obtain an image; a thermal imaging camera that thermally tracks and detects a person's face to obtain a thermal image; an expression means for outputting preset warning information to the outside; The thermal image obtained by the thermal imaging camera is analyzed to measure the temperature of the occupant to detect whether the occupant has a fever, and at the same time, the image acquired by the camera is analyzed to detect whether the occupant is wearing a mask, and the occupant's fever is detected. It is to further include a control unit for outputting warning information through the expression means when this is detected or the person entering and leaving the self-mask does not wear.

In addition, in the present invention, the quarantine system further includes gate devices in which the bracket of the access detector is installed on the upper surface, and the control unit 1) opens the access door with the gate device when the person wearing a mask is not a heat generator It is preferable to output a control signal to close the access door with the gate device when the person entering or leaving is a heat generator or not wearing a mask.

In addition, in the present invention, the control unit includes a memory for storing a feature detection algorithm using an image as an input value and the presence or absence of a mask object of an occupant as an output value; an object recognition and facial pattern feature detection unit that analyzes the image acquired by the camera using the object detection algorithm and outputs object detection data indicating the presence or absence of a mask object; a thermal image analyzer analyzing the thermal image acquired by the thermal imaging camera; After measuring the body temperature (℃) of the visitor through the analysis data detected by the thermal image analysis unit, the measured body temperature (℃) is compared with a preset reference value. However, if the body temperature (°C) is higher than the reference value, a fever determination unit that determines that the person entering the room is a fever; a mask wearing determination unit for determining whether an occupant is wearing a mask through the object detection data detected by the object recognition and facial pattern feature detection unit; When it is determined by the fever determination unit that the entrant is a non-heater and the entrant is determined to be a mask wearer by the mask wearing determination unit, it is determined to allow the entrant, but the fever determination unit determines that the entrant is a person with a fever or an entrance decision unit for determining whether to control the entrance or exit of the entrant when it is determined that the entrant is not wearing a mask by the mask-wearing determination unit; a control signal generation unit for generating a control signal according to the data determined by the entry/exit determination unit; Preferably, the control module further includes a control module for transmitting the control signal generated by the control signal generator to the corresponding gate device.

In addition, in the present invention, the feature detection algorithm uses an image as an input value, and outputs not only object detection data indicating the presence or absence of a mask object of the occupant but also facial pattern feature information, which is feature information of the occupant's face, and the memory Registration DBs in which user information and facial pattern characteristic information are matched are preset and stored, and the object recognition and facial pattern characteristic detection unit analyzes the image obtained by the camera using the object detection algorithm to determine the presence or absence of a mask object Detects object detection data indicating and the facial pattern characteristic information of the visitor, and the control unit compares and compares the facial pattern characteristic information detected by the object recognition and facial pattern characteristic detection unit with pre-registered registration DBs to allow the entrance The user authentication unit further includes a user authentication unit for determining whether the user has been used, wherein the entry/exit determination unit determines whether an entrant is a non-fever by the fever determination unit, and at the same time determines that the entrant is a mask wearer by the mask wearing determination unit When it is authenticated by the department, it is decided to allow the entrance of the occupant, but the entrant is determined to be a fever by the fever determination unit, or the entrant is determined to be a non-masked person by the mask wearing determination unit, or the user authentication unit In case of failure, it is desirable to decide to control the access of the entrants.

In addition, in the present invention, preset sectors, categories that are sub-components constituting each sector, and authentication order information that is an order of the sectors are preset and stored in the memory, and the user authentication unit performs authentication when authentication succeeds. Stored in the memory for generating authentication information including the processed user information, the authentication required time (t), which is the time required for the authentication process, and category information for each sector corresponding to the time when authentication is performed, and the second user authentication The unit includes: a registration DB sorting unit that is executed every second preset period (T2) to align the registration DBs; Further comprising an authentication processing unit that performs authentication by sequentially comparing and collating the facial pattern characteristic information detected by the object recognition and facial pattern characteristic detection unit, and the registration DB, which is a registration DB arranged by the registration DB alignment unit, respectively, and , the registration DB sorting unit update module for substituting the content of the current state to the category of each sector in a preset authentication order; an access status information generating module for generating access status information including the number of times of authentication (N) for each category of each sector of previously registered users by using the authentication information; a matching data generation module for generating matching data by matching the access status information generated by the access status information generation module with pre-registered registration DBs based on user information; After giving weights proportional to the order for each sector according to the latest authentication order, which is the authentication order in which the content of the current state is applied by the update module, the latest authentication order for each user using the matching data generated by the matching data generation module At the same time, the number of authentications in the same category as the category content of each sector of a weighted score calculation module for calculating a weighted score for each user; It is preferable to include a sorting DB generating module that sorts the registered oils according to the size of the weighted score calculated by the weighted score calculation module to generate a sorting DB, and then stores the sorting DB in the memory.

According to the present invention having the above problems and solutions, by using a camera and a thermal imaging camera to double detect a person with a fever and not wearing a mask, and at the same time, when an event occurs, warning information is expressed to the outside to prevent viruses and By effectively preventing the transmission of infectious diseases such as bacteria, it is possible to maximize the heat dissipation effect.

In addition, according to the present invention, the facial pattern features of the occupant are detected by analyzing the image obtained by shooting the camera. As it is configured to perform user authentication compared to DBs, authentication is possible without requiring a separate mask to be removed to perform user authentication. can be effectively prevented.

In addition, according to the present invention, as components such as a camera, a thermal imaging camera, a speaker, and an LED light are integrally manufactured, miniaturization is possible, and manufacturing and operating costs can be reduced.

Also, according to the present invention, when user authentication is performed, the registration DBs are configured to be compared and compared in response to the user's access status, thereby remarkably increasing the authentication speed and preventing unnecessary computational processing.

1 is a diagram showing the construction of a fever detection system disclosed in Korean Patent Registration No. 10-2091121 (Title of the Invention: Arrivals Fever Detection System).
2 is a block diagram showing the quarantine system according to an embodiment of the present invention.
3 is an exemplary view of FIG. 2 .
FIG. 4 is a block diagram illustrating the non-contact access detector of FIG. 2 .
FIG. 5 is an exemplary view of FIG. 4 .
FIG. 6 is another exemplary view of FIG. 4 .
FIG. 7 is a block diagram illustrating the photographing device of FIG. 4 .
FIG. 8 is an exemplary view of FIG. 7 .
FIG. 9 is another exemplary view of FIG. 7 .
FIG. 10 is an exemplary view illustrating an image obtained by photographing the camera of FIG. 7 .
11 is an exemplary diagram illustrating a thermal image obtained by thermal-detection and tracking of the thermal imaging camera of FIG. 7 .
12 is a block diagram illustrating a control unit of FIG. 7 .
13 is a block diagram illustrating the object recognition and facial pattern feature detection unit of FIG. 12 .
14 is a block diagram illustrating a second embodiment of the control unit of FIG. 12 .
15 is a block diagram illustrating a second user authentication unit of FIG. 14 .
16 is a block diagram illustrating the registration DB alignment unit of FIG. 15 .
17 is a block diagram illustrating an authentication procedure setting unit of FIG. 15 .

Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

FIG. 2 is a block diagram showing a quarantine system according to an embodiment of the present invention, and FIG. 3 is an exemplary view of FIG. 2 .

The quarantine system 1 according to an embodiment of the present invention 1) maximizes the prevention rate and quarantine efficiency of infectious diseases by double detecting and cracking down on those who have a fever or not wearing a mask among the entrants, and 2) Authenticates the occupants using facial recognition When user authentication is performed by adding a function, it is configured to enable facial authentication without the user having to remove a separate mask, simplifying unnecessary authentication procedures and increasing convenience of use. This is to maximize the authentication speed by giving priority to previously registered registration information (registration DB), which is to be compared, and then composing a matching operation during facial authentication according to the assigned priority.

In addition, as shown in FIGS. 2 and 3, the quarantine system 1 of the present invention is installed at each of the entrances 930 of an indoor space S such as an airport, various buildings, government offices, hospitals, etc. It is installed in each of the gate devices (5-1), ..., (5-N) and gate devices (5-1), ..., (5-N) that open and close the door according to the authentication status. Non-contact access detector (3-1), ..., (3) that captures and thermally detects the face of an occupant and acquires an image (visible light image) and thermal image -N).

At this time, in the present invention, for convenience of explanation, each of the non-contact access detectors 3-1, ..., (3-N) performs image and thermal image analysis, fever determination, mask wearing determination, user authentication, etc. Although it has been described as an example that the processing is performed by itself, it is natural that this operation processing may be made in a controller or a local server installed in the corresponding indoor space (S), or may be configured in a separate central management server.

Gate devices (5-1), ..., (5-N) are respectively installed at the entrances (930) of the indoor space (S) to control the entrance according to the control of the non-contact type access detector (3) for access control. It opens and closes to control and regulate the entry and exit of entrants.

Since the structures and shapes of these gate devices 5-1, ..., (5-N) are already widely used technologies, various structures and shapes of known gate devices can be applied.

4 is a block diagram showing the non-contact access detector of FIG. 2 , FIG. 5 is an exemplary view showing a reduced state of the vertical arms of the housing of FIG. 4 , and FIG. 6 is an example showing an expanded state of the vertical arms of FIG. 5 It is also

The non-contact access detector 3 rotates on the upper portion of the housing 32 and the housing 32 comprising a fixed plate 321 , vertical arms 323 and a bracket 325 as shown in FIGS. 4 to 6 . It consists of a device 30 for photographing that is possibly installed.

The housing 32 includes a fixing plate 321 , vertical arms 323 , and a bracket 325 .

The fixing plate 321 is formed of a disk-shaped plate, and the lower surface is seated on the ground or a structure.

At this time, a power supply means such as a battery is installed inside the fixing plate 321 .

The vertical arms 323 are vertically installed on the upper surface of the fixing plate 321 , and are coupled to be extendable/retractable in the height direction. In this case, a configuration in which a plurality of pipes or frames are coupled to be reduced and extended is already widely used, and thus a detailed description thereof will be omitted.

In addition, a bracket 325 is rotatably coupled to the upper end of the upper vertical arm among the vertical arms 323 .

The bracket 325 is rotatably installed at the upper end of the upper vertical arm, and the photographing device 30 is seated therein.

7 is a block diagram illustrating the photographing device of FIG. 4 , FIG. 8 is an exemplary view of FIG. 7 , and FIG. 9 is another exemplary view of FIG. 7 .

The photographing device 30 is installed so as to be able to adjust the front and rear angles by being seated on the bracket 325 of the housing 32 .

In addition, as shown in FIGS. 7 to 9 , the device 30 for photographing includes a case 32 having a display panel 321 installed on the front side, and a case 32 installed on the front side of the case 32 to capture an image by photographing entrants. The camera 31, the TOF (Time Of Flight) sensor 33 installed in the front of the case 32, and the heat installed in the upper part of the case 32 to thermally detect and track the occupants to obtain a thermal image An image camera 34, a speaker 35 installed in the lower part of the case 32 to output guidance and warning messages, an LED part 36 installed in the front of the case 32 to turn on lighting, and the case It is installed inside the 32 and consists of a control unit 300 for managing and controlling the operation of these constituent means (31), (33), (34), (35), (36).

At this time, in the present invention, for convenience of explanation, the structure and shape of the non-contact access detector 3 and the photographing device 30 have been described as being configured as shown in FIGS. 5 and 6 as an example, but the non-contact access detector 3 and the photographing device The structure and shape of (30) is not limited thereto, and may be configured in various structures and shapes.

FIG. 10 is an exemplary diagram illustrating an image obtained by photographing the camera of FIG. 7 , and FIG. 11 is an exemplary diagram illustrating a thermal image obtained by thermal-detection and tracking of the thermal imaging camera of FIG. 7 .

The camera 31 is installed on the front of the case 32 , obtains the image 910 of FIG. 7 by photographing the face of an entrant, and outputs the obtained image 910 to the controller 300 .

At this time, the camera 31 is installed with a Pan-Tilt-Zoom (PTZ) set so that the face of the person who is located at the entrance door of the gate device 5 is photographed.

The thermal imaging camera 34 is installed on the upper part of the case 32 to thermally detect and track the face of the entrant to obtain a thermal image 920 , and output the acquired thermal image 920 to the controller 300 . .

At this time, the thermal imaging camera 34 is installed with a PTZ (Pan-Tilt-Zoom) set so that the face of the occupant located at the entrance door of the gate device 5 is heat-detected.

The TOF sensor 33 includes light transmitters (not shown) that radiately output N optical signals, and light receivers (not shown) that are emitted by the light transmitters and receive reflected waves, and transmit and receive optical signals The information is output to the control unit 300 .

That is, since the optical signal information (hereinafter referred to as TOF information) transmitted and received by the TOF sensor 33 of the present invention has different values depending on the size and depth of the face of the occupant, the controller 300 controls the TOF sensor ( 33) can be used for facial pattern feature information by analyzing the TOF information.

12 is a block diagram illustrating a control unit of FIG. 7 .

The control unit 300 of FIG. 12 is a microprocessor installed inside the case 32 to manage and control the operation of the constituent means 31 , 33 , 34 , 35 , and 36 .

In addition, as shown in FIG. 12 , the control unit 300 includes a control module 301 , a memory 302 , a communication interface unit 303 , an object recognition and facial pattern feature detection unit 304 , and a thermal image analysis unit 305 . ), fever determination unit 306, mask wearing determination unit 307, user authentication unit 308, access determination unit 309, opening/closing control signal generation unit 310, warning information output unit 311 , a display unit 312 , an electronic access list management unit 313 , and an untack manager management unit 314 .

The control module 301 is an OS (Operating System) of the control unit 300, and is a control target 302, 303, 304, 305, 306, 307, 308, 309 ), (310), (311), (312), (313), and (314) are managed and controlled.

In addition, the control module 301 inputs the image and TOF information input from the camera 31 and the TOF sensor 33 through the communication interface unit 303 to the object recognition and facial pattern feature detection unit 304 .

In addition, the control module 301 inputs the thermal image input from the thermal imaging camera 34 to the thermal image analysis unit 305 .

In addition, the control module 301 inputs the result data detected by the fever determination unit 306 , the mask wearing determination unit 307 , and the user authentication unit 308 to the entry/exit determination unit 309 .

The memory 302 temporarily stores images acquired by the camera 31 , the thermal imaging camera 34 , and the TOF sensor 33 , the thermal image and TOF information.

In addition, a preset feature detection algorithm is stored in the memory 302 .

In this case, the feature detection algorithm is a normal machine learning algorithm that can be learned, and in detail, it uses the image and TOF information as input values, and analyzes the input values using the learned model, such as face, mask, glasses, etc. It is an algorithm for detecting an object of , and outputting facial pattern feature information, which is a pattern feature of the upper region centered on the eye of the detected face object.

In addition, pre-registered registration information (registration DB) is stored in the memory 302 . In this case, the registration DB means information on the facial pattern characteristics of the allowed access.

The communication interface unit 303 inputs and outputs data to and from the camera 31 , the thermal imaging camera 34 , and the TOF sensor 33 .

In addition, the communication interface unit 303 transmits/receives data to and from an external server and device.

13 is a block diagram illustrating the object recognition and facial pattern feature detection unit of FIG. 12 .

As shown in FIG. 13 , the object recognition and facial pattern feature detection unit 304 includes an image input module 341 , a TOF information input module 342 , and an analysis module 343 .

The image input module 341 receives an image obtained by photographing the camera 31 through the communication interface unit 303 under the control of the control module 301 .

The TOF information input module 342 receives TOF information input from the TOF sensor 33 through the communication interface unit 303 under the control of the control module 301 .

The analysis module 343 uses a preset feature detection algorithm to analyze the input image and TOF information to detect an object such as a face, mask, and glasses, and at the same time detect facial pattern feature information that is a feature of the detected facial object do.

In this case, the feature detection algorithm is a normal machine learning algorithm that can be learned, and in detail, it uses the image and TOF information as input values, and analyzes the input values using the learned model, such as face, mask, glasses, etc. It is an algorithm for detecting an object of , and outputting facial pattern feature information, which is a pattern feature of the upper region centered on the eye of the detected face object.

That is, the present invention does not detect the pattern features of the entire face using the feature detection algorithm, but is configured to detect the pattern features of the upper region centered on the eyes of the face object, so that the occupant wears and separates the mask separately. Because user authentication is possible without performing user authentication, authentication is possible without the need for a separate mask to be removed to perform user authentication, thereby enhancing the convenience and efficiency of the authentication process and effectively preventing unnecessary authentication time delay due to wearing and removing the mask. have the advantage of being able to

In addition, the control module 301 inputs the object detection data detected by the analysis module 343 to the mask wearing determination unit 307, and the facial pattern characteristic information detected by the analysis module 343 to the user authentication unit ( 308).

The thermal image analysis unit 305 analyzes a thermal image obtained by thermal-detection and tracking of the thermal imaging camera 34 .

At this time, the analysis data detected by the thermal image analysis unit 305 is input to the fever determination unit 306 under the control of the control module 301 .

The fever determination unit 306 measures the body temperature (°C) of the occupant through the analysis data detected by the thermal image analysis unit 305 .

In addition, the fever determination unit 306 compares the measured body temperature (°C) with a preset reference value and, if the body temperature (°C) is less than the reference value, judges the person entering as a non-fever, but if the body temperature (°C) is higher than the reference value, The entrants are judged to be feverish.

In addition, the fever determination unit 306 outputs the detected determination data to the entry/exit determination unit 309 .

At this time, the control module 301 operates the warning information output unit 311 when it is determined by the fever determination unit 306 that a person is a person with a fever.

The mask wearing determination unit 307 utilizes and analyzes the object detection data input from the object recognition and facial pattern feature detection unit 304 to determine whether the entrant is wearing a mask.

Also, if the mask object is included in the object detection data, the mask wearing determination unit 307 determines that the person wearing the mask is wearing a mask, but if the object detection data does not include the mask object, it is determined that the person is not wearing the mask do.

Also, the mask wearing determination unit 307 outputs the detected determination data to the entry/exit determination unit 309 .

At this time, the control module 301 operates the warning information output unit 311 when it is determined by the mask wearing determination unit 307 that the entrant is not wearing a mask.

The user authentication unit 308 verifies and compares the facial pattern characteristic information input from the object recognition and facial pattern characteristic detection unit 304 with pre-registered registration DBs to authenticate whether the person is an authorized user.

In addition, the user authentication unit 308 determines that the user authentication is successful if the user is an authorized user, but determines that the user authentication has failed if the user is not an authorized user.

In addition, the user authentication unit 308 outputs the detected authentication result to the access decision unit 309 .

At this time, when it is determined by the user authentication unit 308 that the user authentication of the visitor has failed, the control module 301 operates the warning information output unit 311 .

The entry/exit determination unit 309 is configured to 1) determine whether a person has a fever by the fever determination unit 306 or 2) determine whether the person wearing a mask is not wearing a mask, or 3) a user authentication unit 308 ), when the user authentication of the entrant fails, it is decided to control the entrant's access, and the detected decision data is output to the opening/closing control signal generating unit 310 .

In addition, the entry/exit determination unit 309 1) determines whether a person has a fever by the fever determination unit 306, and at the same time 2) determines whether the person wearing a mask is a mask wearer, 3) authenticates the user When the user authentication of the entrant is successful in the unit 308 , it is decided to allow the entry of the entrant, and the detected determination data is output to the opening/closing control signal generating unit 310 .

The opening/closing control signal generating unit 310 generates a control signal for closing the access door 51 of the gate device 5 when it is determined by the access decision unit 309 to control the access of the corresponding person.

In addition, the opening/closing control signal generating unit 310 generates a control signal for opening the entry/exit door 51 of the gate device 5 when it is determined by the entry/exit determination unit 309 to allow the corresponding person to enter or exit.

At this time, when a control signal is generated by the opening/closing control signal generating unit 310 , the control module 301 controls the communication interface unit 303 to transmit the generated control signal to the gate device 5 . ), the access door 51 is opened and closed according to the control of the control module 301 .

The warning information output unit 311 1) outputs the pre-registered first guidance and warning message to the outside through the speaker 35 when it is determined by the fever determination unit 306 that an occupant has a fever.

In addition, the warning information output unit 311 outputs the second registered information and warning message to the outside through the speaker 35 when it is determined by the mask wearing or not determining unit 307 that the person is not wearing a mask. .

In addition, the warning information output unit 311 3) when the user authentication fails by the user authentication unit 308, the previously registered third guide and warning message to the outside through the speaker (35).

The display unit 312 displays an image, a thermal image, and pre-made content through the display panel 321 of the case 32 .

The electronic access list management unit 313 creates a list of persons who have passed through the gate device 5 and stores it in an external security server, so that quarantine management and epidemiological investigation can be easily performed.

In addition, the electronic access list management unit 313 is configured to collect the visitor information by reading the QR code exposed on the smart phone of the visitor, including a QR code reader.

The untack manager management unit 314 manages and controls the untack manager, which is an application that provides the manager with an editing interface such as adding, updating, and deleting information of the registration DB.

At this time, Untack Manager provides a Graphical User Interface (GUI) that allows users to add, edit, and delete data in the registration DB, so that the user can conveniently edit the registration DB through the Untack Manager. be able to do

14 is a block diagram illustrating a second embodiment of the control unit of FIG. 12 .

The second control unit 400 of FIG. 14 is a second embodiment of the control unit of the device 30 for photographing according to the present invention.

In addition, the second control unit 400 includes the communication interface unit 303, the object recognition and facial pattern feature detection unit 304, the thermal image analysis unit 305, and the fever determination unit 306 of the control unit 300 of FIG. ), a mask wearing or not determining unit 307 , an access or not determining unit 309 , an opening/closing control signal generating unit 310 , a warning information output unit 311 , and a display unit 312 .

In addition, the second control unit 400 further includes a second control module 401 , a second memory 402 , and a second user authentication unit 408 .

The second control module 401 performs the same operation as the control module 301 of FIG. 12 described above, but executes the registration DB sorting unit 45 of FIG. 16 every second preset period T2.

In addition, the second control module 401 executes the authentication sequence setting unit 46 of FIG. 15 , which will be described later, every preset first period T1 . At this time, the authentication order information means a preset authentication order of each sector, and the authentication order information set by the authentication order setting unit 46 is utilized in the registration DB sorting unit 45 . In this case, the sector means a type capable of classifying the authentication result, and each sector includes a plurality of categories that are sub-components of the sector.

For example, the sector may be set to 'time zone', 'day of the week', and 'date', and the category of the 'time zone' sector is '00:00 to 02:00', ..., '22:00 to 24'. The category of the 'day' sector may be composed of 'Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun', and The category may be composed of '1 day', '2 days', ..., '30 days', and '31 days'.

Like the memory 302 of FIG. 12 described above, the registration DB (registration information) of pre-registered users is preset and stored in the second memory 402 . At this time, the registration DB is stored in the second memory 402 by converting the user information of the registered user and the facial pattern characteristic information of the user into a DB.

In addition, the authentication information generated by the authentication information generating unit 44 of FIG. 15 to be described later is stored in the second memory 402 . At this time, the access status information includes the authenticated user information, the time required for authentication, and the category contents of each sector for which authentication is performed.

In addition, the second memory 402 stores the sort DB, which is a registration DB sorted by the registration DB sorting unit 45 of FIG. 16 to be described later.

In addition, the authentication procedure information set by the authentication procedure setting unit 46 of FIG. 15 to be described later is stored in the second memory 402 .

15 is a block diagram illustrating a second user authentication unit of FIG. 14 .

As shown in FIG. 15 , the second user authentication unit 408 includes an authentication processing unit 43 , an authentication information generating unit 44 , a registration DB sorting unit 45 , and an authentication order setting unit 46 .

The authentication processing unit 43 is a processor that performs user authentication.

In addition, the authentication processing unit 43 sequentially compares and contrasts the facial pattern characteristic information input from the object recognition and facial pattern characteristic detection unit 304 and the alignment DBs arranged by the registration DB alignment unit 45, respectively, to detect the similarity. do. At this time, since the technique and method for detecting the similarity of a specific pattern of the subject are techniques commonly used in biometric authentication systems, a detailed description thereof will be omitted.

In addition, the authentication processing unit 43 1) If the similarity of the two data is less than a preset threshold, it is determined that the two data do not match, 2) If the similarity of the two data is greater than the threshold, it is determined that the two data are identical, If a sorted DB having a similarity greater than or equal to the threshold is detected, it is determined that authentication is successful and authentication is terminated.

In addition, the authentication result (authentication success or authentication failure) information detected by the authentication processing unit 43 is input to the authentication information generating unit 44 .

When the authentication is successful by the authentication processing unit 43, the authentication information generating unit 44 includes the user information of the corresponding person, the time taken for authentication (t), and the category for each sector in which authentication is performed (authentication date, authentication day, authentication). time) to create authentication information by matching the contents.

The registration DB sorting unit 45 is executed according to the control of the second control module 401 every second preset period T2.

16 is a block diagram illustrating the registration DB alignment unit of FIG. 15 .

As shown in FIG. 16, the registration DB sorting unit 45 includes an update module 451, an access status information generation module 452, a matching data generation module 453, a weighted score calculation module 454, and an alignment DB. It consists of a generation module (455).

The update module 451 substitutes the current state (current time, today's day, today's date) into the categories of each sector in the authentication order set by the authentication order setting unit 46 . At this time, the authentication order in which the current state is substituted will be referred to as the latest authentication order. For example, assuming that the authentication order of the authentication sequence information is 1) time, 2) day, 3) date, and the current status is '10-11', 'Saturday', '14th', the update module (451) substitutes the current status to the authentication sequence information and updates it in the latest authentication order: 1) 10-11 o'clock, 2) Saturday, and 3) 14 days.

The access status information generating module 452 uses the authentication information stored in the second memory 402 to generate access status information including the number of times of authentication (N) for each category of each sector of previously registered users.

For example, access status information includes user information (A), time range (13~14:00: 5 times, 23~24: 12 times), day of the week (Monday: 9 times, Wednesday: 1 time, Friday: 7 times) , days (1 day: 1 time, 2 days: 1 time, 10 days, 5 times, 20 days: 5 times, 30 days: 5 times).

The matching data generation module 453 generates matching data by matching the access status information generated by the access status information generation module 452 with the pre-registered registration DBs based on the user criteria.

The weighted score calculation module 454 assigns a weight to each sector according to the latest authentication order updated by the update module 451, but gives weight only to the category of each sector in the current state. In this case, the weight has a higher value as the authentication order is higher, and has a smaller value as the authentication order is lower.

For example, assuming that the latest authentication order information is 1) 10-11 o'clock, 2) Saturday, and 3) 14 days, the weighted score calculation module 454 determines that the current state of the highest priority 'time' sector is The applied category '10-11' is given the highest weight of 'A', and the next highest weight of 'B' (B<A ), and the lowest weight, 'C' (C<B<A), is given to '14 days', the category to which the current state of the last 'date' sector is applied.

In addition, the weighted score calculation module 454 uses the matching data generated by the matching data generation module 453 to multiply the number of authentications in the same category as the category content of each sector in the latest authentication order for each user by weight, and then sums them up. Thus, a weighted score is calculated for each user in a way that calculates a weighted score. At this time, the category content of each sector in the latest authentication order and the number of authentications in other categories are multiplied by an integer '1' (1<C<B<A).

The sort DB generation module 455 aligns the previously registered registration DBs according to the size of the weighted score calculated by the weighted score calculation module 454 to generate the sort DB. The generated sort DBs are stored in the second memory 402 and used for authentication of the authentication processing unit 43 .

The authentication sequence setting unit 46 is executed according to the control of the second control module 402 every preset first period T1 .

17 is a block diagram illustrating an authentication procedure setting unit of FIG. 15 .

As shown in FIG. 17 , the authentication sequence setting unit 46 includes a determination module 461 and an authentication module 463 for each sequence combination.

The determination module 461 determines whether it is necessary to reset the authentication order.

In addition, the determination module 461 includes a data extraction module 4611 , an average value calculation module 4612 , and a reset or not determination module 4613 .

The data extraction module 4611 searches for authentication information stored in the second memory 402 and extracts authentication information that has been authenticated during the previous period (T-1) from the present.

The average value calculation module 4612 extracts the authentication time t of the authentication information for the previous period (T-1) from the present extracted by the data extraction module 4611 and calculates an average value thereof.

If the average value calculated by the average value calculation module 4612 is greater than or equal to a preset threshold, the reset determination module 4613 determines that resetting of the authentication sequence information is not necessary, but if the average value is less than the threshold, resetting the authentication sequence information deem it necessary In this case, the threshold is defined as the minimum value of the authentication time required to determine the need to reset the priority information.

In addition, if the second control module 401 determines that the reset of the authentication sequence is not necessary by the determination module 461, it does not perform a separate operation and uses the authentication sequence of the previous cycle (T-1) as it is, but resets If it is determined that this is necessary, the authentication module 463 for each sequence combination is executed.

The authentication module 463 for each sequence combination is driven when the determination module 461 determines that the authentication sequence needs to be reset.

The authentication module for each sequence combination 463 includes a test sample extraction module 4631, an authentication sequence combination module 4632, an authentication execution module 4633, a data calculation module for each sequence combination 4634, and an authentication sequence information reset module 4635 ) is made of

The test sample extraction module 4631 searches the facial pattern features of the authenticated users during the previous period (T-1) stored in the second memory 402 and randomly extracts n test samples from among them. That is, the test sample extraction module 4631 randomly extracts facial pattern features of the entrants who have performed user authentication during the previous period (T-1).

The authentication sequence combination module 4632 configures sequence combinations that can be made by a combination of the sequences of sectors. For example, assuming that the sectors are 'day of the week', 'time zone', and 'date', the authentication sequence combination module 4632 performs a) time-day-date, b) time-date-day, c) day- Six sequence combinations such as time-date, d) day-date-time, e) date-time-day, and f) date-day-time can be configured.

The authentication performing module 4633 includes an update module 451 of the registration DB sorting unit 45 described above, an access status information generating module 452, a matching data generating module 453, a weighted score calculation module 454, and sorting. In the same manner as the DB generation module 455 , sorted DBs are generated for each sequence combination. At this time, the current state allows the same value to be applied.

For example, assuming that the sectors are 'day of the week', 'time zone', and 'date', the sequence combination consists of six as described above, and accordingly, the authentication performing module 4633 selects the sort DBs for each sequence combination. be able to create

In addition, the authentication performing module 4633 compares and compares the test patterns extracted by the test sample extraction module 4631 with the generated sorting DBs. In other words, the authentication performing module 4633 performs authentication by comparing each test pattern with six sorting DBs.

The data calculation module 4634 for each sequence combination calculates the authentication time t, which is the time required for each test pattern to authenticate using each sort DB, by the authentication execution module 4633, and then tests for each sort DB. The average value of the authentication time t of the patterns is calculated.

For example, when two test patterns are extracted and six sort DBs for each sequence combination are generated, the data calculation module 4634 for each sequence combination can calculate a total of 12 authentication time t, and sort A total of six average values can be calculated based on the DBs.

At this time, the average values calculated by the data calculation module 4634 for each sequence combination are input to the authentication sequence information resetting module 4635 .

The authentication sequence information resetting module 4635 compares the average values input from the data calculation module 4634 for each sequence combination, and resets the sequence combination having the highest average value as the authentication sequence. At this time, the priority information reset by the authentication sequence information resetting module 4635 is stored in the second memory 402 under the control of the second control module 401 .

As described above, the quarantine system 1 according to an embodiment of the present invention uses the camera 31 and the thermal imaging camera 34 to double detect a person with a fever and a non-mask among the entrants and, at the same time, provide warning information when an event occurs. By exposing it to the outside, it is possible to effectively prevent the transmission of infectious diseases such as viruses and bacteria from the outside, thereby maximizing the heat dissipation effect.

In addition, the quarantine system 1 of the present invention analyzes the image acquired by the camera 31 to detect the facial pattern features of the occupant, but the facial pattern features, which are the pattern features of the upper region centered on the eyes of the face After detection, it is configured to perform user authentication by comparing it with the previously registered registration DBs, so that authentication is possible without requiring an entrant to separately remove a mask to perform user authentication, increasing the convenience and efficiency of the authentication process and wearing a mask at the same time and unnecessary delay in authentication time due to separation can be effectively prevented.

In addition, the quarantine system 1 of the present invention is not only capable of miniaturization, but also manufacturing and It can reduce operating costs.

In addition, the quarantine system 1 of the present invention is configured to compare and compare registered DBs in response to the user's access status during user authentication, thereby significantly increasing the authentication speed and preventing unnecessary computational processing. there is.

1: Quarantine system 3-1, ..., 3-N: Entrance detector
5-1, ..., 5-N: gate device 30: photographing device 31: camera
32: Bracket 33: TOF sensor 34: Thermal imaging camera
35: speaker 36: LED unit 37: control unit
300: control unit 301: control module 302: memory
303: communication interface unit 304: object recognition and facial pattern feature detection unit
305: thermal image analysis unit 306: fever determination unit
307: mask wearing determination unit 308: user authentication unit 309: access determination unit
310: opening/closing control signal generating unit 311: warning information output unit 312: display unit
321: fixed plate 322: caster 323: vertical arm
325: hinge coupling unit 341: image input module 342: TOF information input module
343: analysis module

Claims (5)

In the quarantine system including an access detector installed at the entrance of the indoor space to detect the occupants:
The entrance detector
Brackets;
and a photographing device rotatably coupled to the bracket,
The photographing device is
A camera for acquiring an image by photographing the face of the person entering;
a thermal imaging camera that thermally tracks and detects the face of an occupant and acquires a thermal image;
an expression means for outputting preset warning information to the outside;
The thermal image obtained by the thermal imaging camera is analyzed to measure the temperature of the occupant to detect whether the occupant has a fever, and at the same time, the image acquired by the camera is analyzed to detect whether the occupant is wearing a mask, and the occupant's fever is detected. Prevention system according to claim 1, further comprising a control unit for outputting warning information through the expression means when this is detected or the person entering and leaving the self-mask is not wearing the mask. The method of claim 1, wherein the quarantine system further comprises gate devices installed on the upper surface of the bracket of the access detector,
the control unit
1) When the occupant is not a heat generator but wears a mask, the gate device outputs a control signal to open the door, and 2) controls the gate device to close the door when the occupant is a heat generator or does not wear a mask A quarantine system, characterized in that it outputs a signal. According to claim 2, wherein the control unit
a memory for storing a feature detection algorithm using an image as an input value and determining whether an occupant has a mask object as an output value;
an object recognition and facial pattern feature detection unit that analyzes the image acquired by the camera using the object detection algorithm and outputs object detection data indicating the presence or absence of a mask object;
a thermal image analyzer analyzing the thermal image obtained by the thermal imaging camera;
After measuring the body temperature (℃) of the visitor through the analysis data detected by the thermal image analysis unit, the measured body temperature (℃) is compared with a preset reference value. However, if the body temperature (°C) is higher than the reference value, a fever determination unit that determines the person entering the room as a fever;
a mask wearing determination unit that determines whether an occupant is wearing a mask through the object detection data detected by the object recognition and facial pattern feature detection unit;
When it is determined by the fever determination unit that the entrant is a non-heater and the entrant is determined to be a mask wearer by the mask wearing determination unit, it is decided to allow the entrants, but the fever determination unit determines that the entrant is a person with a fever or an entrance decision unit for determining whether to control the entrance or exit of the entrant when it is determined that the entrant is not wearing a mask by the mask-wearing judging unit;
a control signal generation unit for generating a control signal according to the data determined by the entry/exit determination unit;
The quarantine system further comprising a control module for transmitting the control signal generated by the control signal generator to the corresponding gate device. The method according to claim 3, wherein the feature detection algorithm outputs not only object detection data indicating the presence or absence of a mask object of the occupant but also facial pattern feature information, which is feature information of the occupant's face, using an image as an input value;
In the memory, information of the user allowed to enter and the registration DB in which the facial pattern characteristic information is matched are preset and stored,
The object recognition and facial pattern feature detection unit
Analyze the image acquired by the camera using the object detection algorithm to detect object detection data indicating the presence or absence of a mask object and facial pattern feature information of the person entering and leaving,
the control unit
The object recognition and facial pattern feature detection unit further comprises a user authentication unit for determining whether the user is an authorized user by comparing and comparing the facial pattern feature information detected by the registered DB,
The access decision unit
When it is determined by the fever determination unit that the occupant is a non-heater and the entrant is determined to be a mask wearer by the mask wearing determination unit and authenticated by the user authentication unit, it is determined to allow entry of the entrant, Prevention, characterized in that when the entrant is determined to be a person with a fever by the determination unit, whether the entrant is judged to be not wearing a mask by the mask-wearing determination unit, or when the user authentication unit fails to authenticate, it is determined to control the entrance of the entrant system. 5. The method of claim 4, wherein preset sectors, categories that are sub-components constituting each sector, and authentication order information that is an order of the sectors are preset and stored in the memory;
The user authentication unit
When the authentication is successful, the authentication information is stored in the memory to generate authentication information including the user information, the authentication time required for the authentication process (t), and category information for each sector corresponding to the time when the authentication is made,
The second user authentication unit
a registration DB sorting unit which is executed every second preset period (T2) to align the registration DBs;
Further comprising an authentication processing unit that performs authentication by sequentially comparing and collating the facial pattern characteristic information detected by the object recognition and facial pattern characteristic detection unit, and the registration DB, which is a registration DB arranged by the registration DB alignment unit, respectively, and ,
The registration DB sorting unit
an update module for substituting the content of the current state into the category of each sector in the preset authentication order;
an access status information generation module for generating access status information including the number of times of authentication (N) for each category of each sector of previously registered users by using the authentication information;
a matching data generation module for generating matching data by matching the access status information generated by the access status information generation module with pre-registered registration DBs based on user information;
After assigning a weight proportional to the order to each sector according to the latest authentication order, which is the authentication order in which the content of the current state is applied by the update module, the latest authentication order for each user using the matching data generated by the matching data generation module At the same time, the number of authentications in the same category as the category content of each sector of a weighted score calculation module for calculating a weighted score for each user;
and a sorting DB generating module for aligning the registered oils according to the weighted score calculated by the weighted score calculation module to generate a sorting DB, and then storing the sorting DB in the memory.

Images