KR20230043662A - Moving line tracking system and method for arrival hall of liability to crime travelers - Google Patents

Abstract

The present invention relates to a method for tracking a crime traveler in an arrival hall, which can increase the accuracy of control. The method comprises: a step of comparing prior analysis information on an entrant with a crime traveler database of a traveler information system to find and designate a crime traveler from among entrants; a step of identifying and specifying a crime traveler in comparison to crime traveler information designated through face recognition using image information obtained by a CCTV camera installed in an immigration checkpoint; a step of extracting a facial feature of the crime specified traveler, color and texture feature thereof, and gait feature thereof to generate and store object feature information in real time; a step of extracting facial features of entrants, color and texture feature thereof, and gait feature thereof by using image information obtained by a CCTV camera installed in the arrival hall to generate and store object feature information in real time; a step of detecting the sameness and the similarity between the generated object feature information of the crime traveler and the object feature information of entrants in the arrival hall; a step of recognizing an entrant in the arrival hall, having at least one object feature information with a similarity equal to or higher than a predetermined threshold value, as the crime traveler; and a step of additionally storing the object feature information of the designated crime traveler and providing the same to a customs situation room and customs personnel.

Description

System and method for tracing criminal travelers in the arrival hall {MOVING LINE TRACKING SYSTEM AND METHOD FOR ARRIVAL HALL OF LIABILITY TO CRIME TRAVELERS}

The present invention relates to a system and method for tracking and monitoring a permissive traveler in an arrival hall using an image recognition camera, and more particularly, in an entry hall of an airport or port, an entrant presumed to be a permissive traveler undergoes a CIQ customs clearance procedure from the entry stage. It relates to a system and method for tracking criminal travelers in the arrivals hall, which can be thoroughly observed and monitored by efficiently tracking them along the movement line even if they are changed or disguised to avoid it.

Today, cross-border movement of goods, capital, and services, and the number of domestic and foreign nationals traveling in and out of the country are explosively increasing, making it practically impossible to inspect the baggage of all travelers entering the country at airports or ports.

Accordingly, customs offices around the world pre-emptively provide information on criminal travelers who may commit crimes, such as travelers with prior convictions for weapons or foreign currency smuggling, smuggling, drug trafficking, and terrorism, or travelers who have not properly completed customs procedures in the past. It seeks to improve the efficiency of customs inspection by intensively monitoring and cracking down on criminal travelers when they pass through the country's entry gate.

In recent years, as more and more cases do not voluntarily declare goods that exceed the scope of duty-free duty to customs, when a traveler purchases and brings in expensive goods, the traveler information system (TIS) is provided to track and closely inspect the purchase details and information in advance. : Traveler Information System) has been built and operated.

In other words, customs is required to obtain a passenger list from an airline (hereinafter, including a shipping company) prior to the arrival of an aircraft (hereinafter, including a ship) through the Passenger Name Record (PNR) and the Advance Passenger Information System (APIS). After obtaining the information, the information analyst analyzes the passengers' travel history, destination, use of duty-free shops, etc. passing through quickly.

At this time, by using the traveler information system, the preliminary analysis information of the traveler is compared with the bad traveler DB to designate and determine the subject to be inspected.

In addition, the movement of the inspection subject is tracked using CCTV cameras installed along the movement of the entrants, the customs agent (rover) receives the inspection subject's information through the PDA and monitors the movement, and the inspection marshal (marshal line) It guides entrants passing through the marshal line to the customs inspection point.

However, due to recent changes in the arrival hall environment, such as the unmanned automation of immigration checkpoints and the installation of partitions, there is a limit to accurately identifying the face of a person subject to inspection from CCTV image information, as well as customs, immigration, and It is very difficult to track and monitor if you move in a change of clothes or disguise to avoid the three customs procedures of immigration and quarantine or evade customs officers.

On the other hand, the Intelligent Video Surveillance (IVS) system separates the background area from video information obtained from CCTV, extracts, recognizes, and tracks a moving object, that is, a monitoring target, and monitors the identified object or the object being tracked. Trigger an event appropriate to the situation.

Methods for separating a moving object in such an intelligent video surveillance system include a Gaussian mixture model (GMM), background subtraction, and the like.

Methods for identifying moving objects in an intelligent video surveillance system include contour-based classification and motion-based classification methods. In particular, contour-based classification identifies data sets of points, box-shaped shapes, contours, and blobs from extracted moving objects. classify using

Methods of using color information of moving objects in an intelligent video surveillance system include Mean Shift and CAMshift algorithms. These algorithms extract objects at high speed using the color density distribution of the selected search window.

Methods for tracking moving objects, that is, techniques for finding a moving object identified in a previous image in the current image include region-based tracking, active contour-based tracking, and feature-based tracking. based tracking) and model-based tracking.

Here, the area-based tracking algorithm is a method using changes in the image corresponding to the moving object or structural characteristics (usually, the tracking object is a human and uses the structure of the human head, torso, lower body, etc.). Tracking performance can depend on how you model it.

The active contour-based tracking algorithm is a method of tracking an object by expressing the contour with the boundary and dynamically updating the contour in successive frames. Due to the premise that the contour must be extracted, whether to extract the contour determines the tracking performance. In addition, there is a problem in that automatic tracking is difficult because it is very sensitive to tracking initialization.

Feature-based tracking algorithm is a method of recognizing and tracking an object by matching the features between the features and images through high-level clustering. It is easy to track when a large number of features of a moving object are secured, but it requires a lot of computation to extract feature points. , there are problems such as restrictions (size, color combination, etc.) of moving object images for feature point extraction.

The model-based tracking algorithm is a method of matching after measuring and modeling the structural characteristics of an object in advance, and is mainly a method for recognizing the behavior of an object.

On the other hand, Incheon International Airport Corporation supports the mobility of the transportation vulnerable in the airport, such as arrival and departure halls, and uses AI-based autonomous driving and various cutting-edge information and communication (ICT) services to strengthen various smart services such as temperature checks and information on airport facilities. We have introduced and operated a guide robot that incorporates technology.

In addition, cleaning ability, self-driving function equipped with LiDAR and vision sensors, obstacle avoidance technology, as well as maps of areas requiring cleaning within the airport are stored in the database to recognize the surrounding situation and move freely inside the building on its own. It has introduced and operated an unmanned self-driving cleaning robot that cleans and a service robot that carries luggage while chasing tourists.

The background art or prior art described above herein is information possessed by the present inventor or acquired in the process of deriving the present invention, and is only intended to help understand the technical significance of the present invention, prior to the filing of the present invention, the technology to which this invention belongs It should be noted that it does not mean widely known technologies in the field.

KR 10-2015-0094074 A(2015.08.19) KR 10-2018-0040255 A(2018.04.20) KR 10-1808587 B1(2017.12.07)

Therefore, the present inventor comprehensively considers all the above matters and at the same time, in order to solve the technical limitations and problems of the existing bad traveler tracking system in the arrival hall, the bad traveler who passed the immigration checkpoint is placed in the arrival hall to avoid the CIQ customs clearance procedure. Even if you are disguised or change clothes, such as wearing heavy makeup, glasses, and masks, you can efficiently track along the movement line and thoroughly observe and continuously monitor it. The present invention was created as a result of continuous research with painstaking efforts to develop a system and method for tracking criminal travelers in the arrival hall of the structure.

Therefore, the technical problem and object to be solved by the present invention is to provide a system and method for tracking bad travelers in the arrival hall, which can continuously track and monitor the movements of bad travelers in the arrival hall even if they change clothes or disguise.

Here, the technical problems and objects to be solved by the present invention are not limited to the technical problems and objects mentioned above, and other technical problems and objects that are not mentioned are the general knowledge in the technical field to which the present invention belongs from the description below. Those who have it will be able to clearly understand.

Specific means according to an embodiment of the present invention for effectively achieving a specific technical purpose while embodying a new idea for solving the technical problem of the present invention as described above are prior analysis information and traveler information for the entrant. A prime traveler information input unit that compares the system's bad traveler database and receives bad traveler information among entrants, a first image information input unit that receives image information from a plurality of CCTV cameras installed at the immigration counter, and the first image information input A bad traveler detection unit that identifies the entrant's object from the video information input through the bad tourist information input unit and compares it with the good bad traveler information input through the bad tourist information input unit to detect the good bad traveler; (texture) First object feature information extraction unit that generates and stores object feature information in real time by extracting features and gait characteristics, second image information input unit that receives image information from a plurality of CCTV cameras installed in the arrival hall, A second object feature information extraction unit that generates and stores object feature information in real time by extracting facial features, color and texture features, and gait features of entrants from the image information input through the second image information input unit, and the first object feature A first similarity detection unit that detects the identity and similarity between the object characteristic information of the criminal traveler generated by the information extraction unit and the object characteristic information of the entrants in the arrival hall generated by the second object characteristic information extraction unit; 1. The first high-risk traveler identification unit that recognizes an entrant whose similarity value detected by the similarity detection unit is higher than a predetermined threshold and identifies the entrant as a high-risk traveler, a plurality of CCTV cameras installed at the immigration checkpoint and a plurality of CCTV cameras installed in the arrival hall and a storage unit for storing identification information and arrangement information for each of the plurality of self-driving robots disposed in the arrival hall, object characteristic information and image information of a specified prime traveler, and object characteristic information and image information of a specified prime traveler. A transmission unit provided to the customs control room, customs rover, and marshal, a control unit that controls the functions of the self-driving robots (R) and the overall operation of the system, and wireless communication with the self-driving robots, respectively. A data communication unit that communicates data with the data communication unit, extracts facial features, color and texture characteristics, and gait characteristics of entrants from image information obtained by receiving input through the data communication unit and photographing with the cameras of the self-driving robots to generate object feature information in real time Identity and similarity between the object characteristic information of the prime traveler specified in the first prime traveler specification unit and the object characteristic information of the entrants in the arrival hall generated by the third object characteristic information extraction unit. a second similarity detection unit that detects a second similarity detection unit, a second traveler specification unit that recognizes an entrant whose similarity value detected by the second similarity detection unit is equal to or higher than a predetermined threshold, and re-specifies them as a frequent traveler, and image information received from the self-driving robots. and a location recognition unit that recognizes each location in real time based on GPS, a GIS display unit that displays the location information of the self-driving robots on the display unit, and an image containing a rogue traveler captured by the camera of the self-driving robot Suggested is a system for tracking criminal travelers in an arrival hall, characterized in that it employs an event alarm unit that outputs information in real time to the display unit.

Accordingly, the present invention can not only track a criminal traveler who has passed through the immigration checkpoint at an airport or port along the movement line even if he or she changes clothes or wears heavy makeup, glasses, and a mask in the arrival hall to avoid the CIQ customs clearance procedure. Even if a high-risk traveler hides or is located in a blind spot in the arrival hall outside the shooting range of a fixed CCTV camera, the self-driving robot can be used to collect and analyze feature information of objects photographed from various angles at a height of about a person's height, making it more efficient. By detecting and tracking criminal travelers, it is possible to thoroughly observe and monitor them.

In addition, in a preferred embodiment of the present invention, the control unit receives image information from the self-driving robots, uses the received image information to determine the location of the re-specified criminal traveler, Based on the relative location information corresponding to the distance between the criminal traveler and a specific place in the arrival hall, if the relative location information is out of the standard range, the moving direction of the self-driving robot (R) and the nearest other autonomous robot is determined. It is controlled to change to the specific place, and the location of the re-specified criminal traveler can be reconfirmed using image information received from the other self-driving robot.

And specific means according to another embodiment of the present invention, (a) comparing preliminary analysis information on the entrant with the bad traveler database of the tourist information system to search for and designate bad travelers among the entrants, (b) ) Recognizing the face of an entrant passing through the immigration checkpoint using the image information obtained by shooting a plurality of CCTV cameras installed at the immigration checkpoint, and identifying and detecting the picky traveler by comparing it with designated bad traveler information, (c) detection Creating and storing object feature information in real time by extracting the facial features, color and texture features, and gait features of the high-risk traveler, Creating and storing object feature information in real time by extracting features, color and texture features, and gait features, (e) detecting the degree of identity and similarity between the created object feature information of bad travelers and the object feature information of entrants in the arrival hall (f) recognizing an entrant whose similarity in at least one of the object characteristic information of the entrant in the arrival hall is extracted above a predetermined threshold and specifying the entrant as the pertinent bad traveler; (g) the object characteristic information of the specified bad traveler and additionally storing the image information and providing it to the customs control room, customs rover, and marshal, (h) the facial features and color of the entrants from the image information obtained by the self-driving robots in the arrival hall filming with cameras. And generating and storing object feature information in real time by extracting texture features and gait features, (i) detecting the degree of identity and similarity between the object feature information of the entrants in the created arrival hall and the object feature information of the already specified high-risk traveler (j) Recognizing and re-identifying an entrant whose similarity in at least one of object feature information of the entrant in the arrival hall is higher than a predetermined threshold and (k) image information received from self-driving robots and recognizing each location in real time based on GPS and displaying the location information on a display unit.

As a result, the present invention efficiently tracks and thoroughly observes criminal travelers who have passed through the immigration checkpoint at an airport or port, even if they change or change their clothes, such as wearing heavy makeup, glasses, and masks, in the arrival hall to avoid CIQ customs clearance procedures. and can be monitored.

In addition, in a preferred embodiment of the present invention, in the step (j), the location of the re-specified crime traveler is confirmed using the image information received from the self-driving robots, and the location of the re-specified crime traveler and Based on the relative location information corresponding to the distance between specific places in the arrival hall, if the relative location information is out of the standard range, to change the moving direction of the self-driving robot and another autonomous robot closest to the specific place control, and the location of the re-specified criminal traveler can be reconfirmed using the image information received from the other self-driving robot.

In addition, in a preferred embodiment of the present invention, in step (e), feature point matching is performed to find similar feature points by comparing the extracted facial feature image of the criminal traveler with the feature point descriptor extracted from the facial feature image of the entrants in the arrival hall. Identity and similarity of facial features can be detected.

In addition, in a preferred embodiment of the present invention, in step (e), a color histogram is generated from the extracted color feature images of criminal travelers and color feature images of entrants in the arrival hall, and main image features are extracted. and the sameness and similarity of the color features may be detected only when the color features are equal to or greater than a predetermined threshold value by comparing each other.

In addition, in a preferred embodiment of the present invention, in the step (e), after image segmentation of the extracted texture feature image of the criminal traveler and the texture feature image of the entrants in the arrival hall, MBR (Minimum Boundary Rectangle), apply wavelet transform to the MBR of objects included in each image to generate texture features, and detect their identity and similarity.

In addition, in a preferred embodiment of the present invention, in step (e), a binary silhouette video sequence is obtained from image information obtained by shooting by a plurality of CCTV cameras installed at the immigration checkpoint, and then normalized from the silhouette video sequence. Obtaining a silhouette image, extracting identifiable gait feature information from the normalized silhouette image, and comparing the extracted gait feature information with the gait feature information of the image information obtained by shooting a plurality of CCTV cameras installed in the arrival hall The identity and similarity of features can be detected.

According to the technical idea and embodiment of the present invention, which is based on a unique solution to solve the above technical problems, a bad traveler who has passed through the immigration checkpoint of an airport or port can Even if you are disguised, such as wearing heavy makeup, glasses, or a mask, or change clothes, you can respond and track more efficiently along the movement line, thoroughly observe and continuously monitor, which will increase the accuracy of customs control as well as improve the efficiency of customs work. can

Moreover, by using self-driving robots such as guide robots or cleaning robots operated in the arrival hall, the location and movement of bad travelers are checked in real time, so that when the density of immigrants is high in the arrival hall, inside the bathroom or duty free shop, behind a pillar or sculpture Even if a high-risk traveler hides or is located in a blind spot in the arrival hall outside the shooting range of a fixed CCTV camera, it is possible to collect and analyze characteristic information of objects photographed from various angles at a height of about a person's height, so reliability and tracking of high-risk traveler detection efficiency can be improved.

Here, the effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the description of the claims.

1 is a block diagram schematically illustrating a system for tracking criminal travelers in an arrival hall according to an embodiment of the present invention.
2 is a flow chart illustrating a method for tracing a criminal traveler in an arrival hall according to an embodiment of the present invention.

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

Prior to this, the terms to be described below are defined in consideration of functions in the present invention, which specifies that they should be interpreted as concepts consistent with the technical spirit of the present invention and meanings commonly used or commonly recognized in the art.

In addition, if it is determined that a detailed description of a known function or configuration related to the present invention may obscure the gist of the present invention, the detailed description will be omitted.

The accompanying drawings may be exaggerated or simplified in part to help explain and understand the configuration and operation of the technology, and each component on the drawings does not exactly match the actual size and shape. reveal

In addition, in this specification, the term and / or is meant to include a combination of a plurality of related items described or any item among a plurality of related items described, and when a certain part includes a certain component, this is particularly the opposite. Unless otherwise stated, it means that other components may be further included rather than excluding other components.

That is, terms such as 'include' or 'have' described in this specification mean that a feature, number, step, operation, component, part, or combination thereof exists, but one or the It should be understood that the above does not preclude the possibility of the presence or addition of other features or numbers, step operating components, parts, or combinations thereof.

In addition, each step may occur in a different order from the specified order unless a specific order is clearly described in context. That is, each step may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

And the meaning of the terms "unit" and "unit" used in the present invention means a module type that plays a role or unit that processes at least one function or a certain operation for the purpose of the system, which is hardware or software or It can be implemented as a means through a combination of hardware and software, or as a device or assembly capable of performing an independent operation.

In addition, the meaning of the term "module" used in the present invention may mean a unit including one or a combination of two or more of hardware, software, or firmware, and may also mean a unit, logic It can be used interchangeably with terms such as logic, logical block, component, circuit, etc., and can be a minimum unit or part of an integrated component, It may be a minimum unit or part thereof that performs one or more functions, and may be implemented mechanically or electronically.

In addition, terms such as top, bottom, top, bottom, or top, bottom, top, bottom, front and rear, left and right are used for convenience to distinguish the relative positions of each component. For example, an upper part in the drawing may be named or referred to as an upper part, a lower part may be named or referred to as a lower part, a longitudinal direction may be named or referred to as a front-back direction, and a width direction may be named or referred to as a left-right direction.

Also, terms such as first and second may be used to describe various components. That is, terms such as first and second may be used for the purpose of distinguishing only one element from another element. For example, a first component may be referred to as a second component without departing from the scope of the present invention, and the second component may also be referred to as a first component.

Referring to FIG. 1, the system for tracking criminal travelers in the arrival hall according to an embodiment of the present invention includes a plurality of CCTV cameras (C1) installed in the immigration checkpoint so as not to overlap, and a plurality of CCTV cameras installed in the arrival hall so as not to overlap each other. (C2) can receive the video information obtained by filming and the video information obtained by filming by the plurality of self-driving robots (R) disposed in the arrival hall with cameras and display them on the screen.

In addition, the criminal traveler tracking system in the arrival hall according to an embodiment of the present invention is connected to a plurality of CCTV cameras (C1) (C2) through a wireless or wired network, respectively, and a plurality of self-driving robots (R) and a wireless network. Since data communication is possible through the , images taken from each of them can be transmitted in real time or data can be transmitted or received.

In addition, various types of information such as identification information and arrangement information for each of the plurality of CCTV cameras C1 and C2 and the plurality of self-driving robots R may be registered and stored.

Here, the information for each CCTV camera ((C1) and (C2) and the information for each self-driving robot (R) may include a unique ID and installation location information assigned to each.

On the other hand, the system for tracking criminal travelers in the arrival hall according to an embodiment of the present invention is implemented as a hardware configuration such as a system device and a server, or can communicate with a plurality of CCTV cameras (C1) (C2) and a plurality of self-driving robots (R). Of course, it may also be implemented in the form of software that is mounted on a device or system that is closely connected and performs the processing of the present invention.

Specifically, the main elements constituting the criminal traveler tracking system in the arrival hall according to an embodiment of the present invention are the criminal tourist information input unit 11, the first image information input unit 12, the criminal traveler detection unit 13, and the first feature. Extraction unit 14, second image information input unit 15, second feature extraction unit 16, first similarity detection unit 17, first offender traveler specific unit 18, storage unit 19, transmission unit (20), control unit 21, data communication unit 22, third feature extraction unit 23, second similarity detection unit 24, second offender traveler specific unit 25, location recognition unit 26, GIS It includes a display unit 27, a display unit 28, an event alarm unit 29 and a manager input unit 30.

Here, each component constituting the criminal traveler tracking system in the arrival hall according to an embodiment of the present invention can be understood as a logically separated component rather than a physically separated component.

That is, each component corresponds to a logical component for realizing the technical idea of the present invention, so even if integrated or separated from each other, if the function performed by the logical configuration of the present invention can be realized, it should be interpreted as falling within the scope of the present invention. Of course, if a component performs the same or similar function, it should be construed as within the scope of the present invention regardless of whether or not the name matches.

The bad tourist information input unit 11 serves to receive information on bad travelers among the entrants by comparing preliminary analysis information on the entrant with the bad traveler database of the tourist information system.

The first image information input unit 12 serves to receive image information transmitted by a plurality of CCTV cameras C1 installed at the immigration checkpoint.

Here, the first image information input unit 12 may provide a network interface function for receiving image information from a plurality of CCTV cameras C1 installed at the immigration checkpoint.

The criminal traveler detection unit 13 image-processes the images captured and transmitted by the plurality of CCTV cameras C1 installed at the immigration checkpoint, identifies the entrant's object from the image information, and enters the criminal tourist information input unit 11. It plays a role in detecting high-risk travelers by comparing them with high-risk traveler information input through

The first feature extraction unit 14 serves to generate and store object feature information in real time by extracting facial features, color and texture features, and gait features of the detected criminal traveler.

That is, the first feature extraction unit 14 recognizes a bad traveler from image information obtained by taking pictures of a plurality of CCTV cameras C1 installed at the immigration checkpoint, learns the recognized bad traveler, and extracts object feature information. .

The second image information input unit 15 serves to receive image information transmitted by a plurality of CCTV cameras C2 installed in the arrival hall.

Here, the second video information input unit 15 may provide a network interface function for receiving video information from a plurality of CCTV cameras C2 installed in the arrival hall.

The second feature extraction unit 16 serves to generate and store object feature information in real time by extracting facial features, color and texture features, and gait features of an entrant in the arrival hall.

That is, the second feature extraction unit 16 extracts object feature information of entrants from image information obtained by taking pictures of a plurality of CCTV cameras C2 installed in the arrival hall.

The first similarity detection unit 17 serves to detect identity and similarity between the generated object characteristic information of the criminal traveler and the object characteristic information of the entrants in the arrival hall.

The first bad traveler specification unit 18 plays a role of recognizing an entrant whose similarity in at least one of the object characteristic information of the entrant in the arrival hall is equal to or higher than a predetermined threshold value and identifying the entrant as the relevant bad traveler.

That is, the first criminal traveler specification unit 18 determines the identity or similarity between the previously detected criminal traveler and the entrant in the arrival hall through similarity analysis.

The storage unit 19 includes various information and data related to the operation of the system, including a plurality of CCTV cameras (C1) installed at the immigration checkpoint, a plurality of CCTV cameras (C2) installed in the arrival hall, and a plurality of autonomous robots ( Each unique number, identification information, angle of view information, arrangement information, driving route information, mission information, location information, status information, operation information, object characteristic information and image information of a specific high-risk traveler are stored for R), and the control unit It serves to provide according to the control of (21).

In addition, in the storage unit 19, identification information and location of a specific CCTV camera C1, C2 or self-driving robot R among the plurality of CCTV cameras C1 and C2 and the self-driving robot R At least one of information, view angle information, status information, arrangement information, placed place information, and image information may exist as matched matching information, which is a specific CCTV camera (C1) (C2) or autonomous robot (R) It can be usefully used to specify a location (or place or area) or to determine the location of a high-income traveler located in that space.

In addition, by storing a 2D or 3D map (map or map information) of the arrival hall in the storage unit 19, the current location of the self-driving robot R or a high-risk traveler can be grasped, or the self-driving robot R can be driven. Routes can be set and used for driving.

Here, the 2D or 3D map of the arrival hall is composed of data for estimating the location based on the image information to determine the current location of the autonomous robot (R) based on the image information received from the autonomous robot (R). It can be.

Here, the storage unit 19 may be implemented with various storage media such as ROM, RAM, hard disk, flash memory, SSD (Solid State Disk), and the like.

On the other hand, the self-driving robot R can drive by driving a plurality of wheels, receive information about the scheduled time to perform the mission through various routes, and can be programmed to perform the mission at that time, and the control unit 21 Various operations and functions such as patrolling, guiding, cleaning, disinfection, transportation, etc. can be performed in the arrival hall by the control of .

In addition, the self-driving robot R may transmit/receive signals including situation information in the arrival hall with the control unit 21 of the system.

In addition, the self-driving robot (R) can monitor the situation at the arrival hall by receiving and integrating the video information captured in each area from the CCTV camera (C2) in the arrival hall.

In addition, the self-driving robot R can receive commands directly from the manager. For example, a command may be directly received from a manager through an input of touching the screen or a voice input, and operations such as patrol, guidance, and cleaning may be performed according to the received command.

In addition, the self-driving robot (R) includes sensors such as a high-resolution camera, a GPS that transmits and receives GPS signals for location tracking and location inquiry, a LiDAR sensor for autonomous driving, and a vision sensor. It is equipped with a camera, so it can recognize the operating environment and judge work through images taken with a camera, and can transmit and receive RF (Radio Frequency) signals.

In addition, the self-driving robot R may include a control module for controlling its operation, obtain state information using sensor information acquired from various types of sensors, or detect (recognize) surrounding environments and objects. data, generate map data, determine travel routes and travel plans, determine responses to interactions, or determine motions.

Here, the self-driving robot R may use sensor information obtained from at least one sensor among lidar, radar, and camera to determine a moving path and a driving plan.

In particular, the self-driving robot R can receive sensor information from external devices to recognize an environment or object in an area where the field of view is obscured or an area over a certain distance, or receive directly recognized information from external devices. .

In addition, the self-driving robot R may recognize surrounding environments and objects using a learning model composed of at least one artificial neural network, and may determine an operation using the recognized surrounding environment information or object information.

In addition, the self-driving robot R determines a movement route and a driving plan using at least one of map data, object information detected from sensor information, or object information obtained from an external device, and controls the driving unit to determine the movement route. and driving according to the driving plan.

Here, the map data may include object identification information about various objects disposed in the space where the autonomous robot R moves. For example, the map data may include object identification information for fixed objects such as walls and doors and movable objects such as potted plants, chairs, and people, and the object identification information may include names, types, distances, locations, and the like. can be included

In addition, the self-driving robot R may perform an operation or drive by controlling a driving unit based on a manager's control/interaction.

For example, the self-driving robot R receives mission information, drives to move to a specific place, receives the location information of the bad traveler from the control unit 21, and connects the bad traveler to the specific place based on the location information. The distance between them can be calculated.

The transmission unit 20 serves to transmit the object characteristic information and image information of the specified criminal traveler to the customs control room, customs rover, and inspection marshal.

The control unit 21 remotely controls the functions of the self-driving robots R and controls the overall operation of the system.

In addition, the control unit 21 image-processes images transmitted from the plurality of CCTV cameras C1 installed at the immigration checkpoint and the plurality of CCTV cameras C2 installed in the arrival hall, and selectively displays them on the display unit 28. Control.

In addition, the control unit 21 receives image information from the self-driving robots R, and uses the received image information to confirm the location of the re-specified criminal traveler.

At this time, the controller 21 determines the relative location information based on the distance between the re-specified criminal traveler and a specific place such as a duty-free shop, toilet, store, quarantine facility, sculpture, aisle, etc. in the arrival hall. If it is out of range, after selecting another self-driving robot (R) closest to the self-driving robot (R) using a GPS signal, etc. By controlling the movement to a specific place (or destination) such as a store, quarantine facility, sculpture, passage, etc., and reconfirming the location of the re-specified criminal traveler using the image information received from the other self-driving robot (R), variable Flexibility and operational efficiency can be maximized to respond efficiently and promptly to the on-site situation.

For example, the control unit 21 determines that the respecified criminal traveler is out of the service area assigned (assigned) to the self-driving robot R when he or she falls more than 5M away from the predetermined sculpture A, and Move other self-driving robots (R) that have been assigned (assigned) to the current location of the prime traveler or the service area that includes the sculpture A in advance, or are located closest to the current sculpture A, toward the re-specified prime traveler or sculpture A can be controlled to

Here, the relative location information refers to a specific place such as a duty-free shop, toilet, store, quarantine facility, sculpture, passage, etc. in the arrival hall and a re-specified bad traveler, assuming that they exist on the same floor. It may mean information about how far away.

Also, the control unit 21 may set a main CCTV camera among a plurality of CCTV cameras C1 and C2 from a manager, and may change the main CCTV camera in the middle according to the manager's selection.

In addition, the control unit 21 uses a plurality of CCTV cameras (C1) (C2) capable of driving PTZF (Pan, Tilt, Zoom, Focal length) to capture a short-range image (zoom image/high-resolution image) of the moving object. each can be controlled.

Further, the control unit 21 can extract and specify identification information of the autonomous robot R based on the identification mark of the image information received from the autonomous robot R, and determine its current location. In addition, the control unit 21 specifies the current location of the self-driving robot (R) that captured the image by using at least one of the time when the image was captured and the location information matched to the autonomous robot (R) that captured the image. can do.

In addition, the control unit 21 may specify coordinate information, such as 3rd floor, area A (1), for the image obtained from the self-driving robot R, and the specified coordinate information is the current location of the self-driving robot R. information can be

At this time, the control unit 21 compares the image information obtained from the autonomous robot R with the map generated by the vision (or visual)-based SLAM (Simultaneous Localization and Mapping) technology to determine the location of the autonomous robot R (or current position) can be estimated. In this case, the control unit 21 specifies an image most similar to the image obtained from the self-driving robot R by using image comparison between the image obtained from the self-driving robot R and the images constituting the already created map, and , It is possible to specify the location information of the self-driving robot (R) in a way of obtaining location information matched to the specified image.

That is, when an image is acquired from the self-driving robot R, the control unit 21 can specify the location (or current location) of the self-driving robot using the acquired image, and a map already stored in the storage unit 19. From the information, it is possible to extract location information (eg, coordinate information) corresponding to the image obtained from the self-driving robot R.

In addition, the control unit 21 may extract the location information matched with the identification information of the CCTV camera C2 or the self-driving robot R that has taken the image from the storage unit 19 and specify it as the location of the perpetrator traveler.

On the other hand, the control unit 21 may receive commands related to the operation of the tracking system through the manager input unit 30 connected to an input device such as a mouse, keyboard, or touch pad, and input or output signals, data, information, etc. can be processed and provided.

The data communication unit 22 serves to perform data communication with each of the self-driving robots R wirelessly.

For example, the data communication unit 22 may receive data such as image information and sensing information from the self-driving robot R in real time and transmit a control command to the self-driving robot R.

Here, the data communication unit 22 is WLAN (Wireless LAN), Wi-Fi (Wireless-Fidelity), Wi-Fi (Wireless Fidelity) Direct, DLNA (Digital Living Network Alliance), WiBro (Wireless Broadband), WiMAX (World Interoperability for Microwave Access), HSDPA (High Speed Downlink Packet Access), HSUPA (High Speed Uplink Packet Access), LTE (Long Term Evolution), LTE-A (Long Term Evolution-Advanced), 5G (5th Generation Mobile Telecommunication), infrared communication (Infrared Data Association; IrDA), UWB (Ultra-Wideband), and Wireless USB (Wireless Universal Serial Bus) technology may be used to communicate with the self-driving robot (R).

The third object feature information extraction unit 23 extracts the facial features, color and texture characteristics, and gait characteristics of the entrants from the image information received through the data communication unit 22 and photographed by the cameras of the self-driving robots (R). It plays a role of generating and storing object feature information in real time.

The second similarity detection unit 24 determines the identity between the object feature information of the prime traveler specified by the first bad traveler specification unit 18 and the object feature information of the entrants in the arrival hall generated by the third object feature information extraction unit 23 and It plays a role in detecting similarity.

The second criminal traveler specification unit 25 recognizes an entrant whose similarity value detected by the second similarity detection unit 24 is equal to or greater than a predetermined threshold value and re-identifies them as prime travelers.

That is, the second bad traveler specification unit 25 determines the identity or similarity between the bad traveler detected in advance and the entrant in the arrival hall through similarity analysis, and re-specifies the bad traveler as the concerned bad traveler.

The position recognition unit 26 serves to recognize the respective positions of the self-driving robots R in real time based on image information and GPS.

The GIS display unit 27 serves to receive real-time location information such as GPS coordinate values of the self-driving robots R and display them on the display unit 28 together with a guide map of the immigration checkpoint and the arrival hall.

That is, the GIS display unit 27 receives a signal transmitted from the self-driving robot R, checks the location information, and displays it on the display unit 28 as a GPS electronic map.

The display unit 28 functions and serves to display various information and data related to the operation of the system on the screen under the control of the control unit 21 .

That is, the display unit 28 may display image information captured and transmitted by the plurality of CCTV cameras C1 and C2 and the plurality of self-driving robots R on the monitor screen.

For example, execution screen information of an application program running in the system or UI (User Interface) and GUI (Graphic User Interface) information according to execution screen information may be displayed.

Here, the display unit 28 includes a liquid crystal display (LCD), a thin film transistor-liquid crystal display (TFT LCD), an organic light-emitting diode (OLED), a flexible display ( It may include at least one of a flexible display), a 3D display, and an e-ink display.

The event alarm unit 29 serves to output on the display unit 28 real-time image information obtained by photographing with the cameras of the self-driving robots R, including the perpetrator traveler.

For example, the event alarm unit 29 generates an alarm with respect to an object of a criminal tourist displayed on the display unit 28, and the current location coordinates of the criminal tourist and a unique number of the self-driving robot R that has photographed the criminal tourist. display

The manager input unit 30 serves to receive management and control commands from the manager.

Here, the type of manager input unit 30 is not particularly limited, and may include at least one of a mechanical input unit and a touch input unit.

For example, the manager input unit 30 consists of a virtual key, soft key, or visual key displayed on a touch screen through software processing, or is disposed on a part other than the touch screen. If it is made of a touch screen, the display unit 28 can perform both a role of outputting information and a role of receiving information.

Meanwhile, referring to FIGS. 1 and 2, a tracking method using the criminal traveler tracking system in the arrival hall according to an embodiment of the present invention will be described.

First, prior analysis information on the entrant is compared with the bad traveler database of the traveler information system to search and designate the bad traveler among the entrants (S101).

Subsequently, the facial recognition of the entrant passing through the immigration checkpoint is performed in real time using the image information obtained from the shooting of the plurality of CCTV cameras (C1) installed in the immigration checkpoint so as not to overlap, and the facial recognition is performed in real time, and the result is compared with the designated risky traveler information. The traveler is identified and detected (S103).

At this time, when image information is received from a plurality of CCTV cameras C1 installed in the immigration checkpoint, an object of interest in the image information may be detected by parsing or analyzing the image information.

Here, the object of interest means a target or object of interest, and may be a person.

In addition, various methods such as an image difference algorithm including methods such as binarization, labeling processing, and edge detection, an optical flow algorithm, and a model matching algorithm may be applied to a method of extracting an object from image information. The detection result can be expressed using Minimum Enclosing Rectangle (MER).

In other words, the width, height, center point coordinates, density information, width/height ratio information, etc. of the MER are defined according to the characteristics of the object, and the defined MER and the characteristics such as the width and height of the object included in the image information are mutually contrasted. A method of detecting/identifying an object of particular interest is used.

Continuously, object feature information is generated and stored in real time by extracting facial features, color and texture features, and gait features of the detected criminal traveler (S105).

In this case, gait recognition extracts features by storing and analyzing the gait information of a person acquired through a camera.

Here, the gait feature can be recognized by deleting the background from the acquired image, extracting the feature based on the outline information of the silhouette, and then comparing it with the feature in the DB.

Alternatively, it may be recognized by a model-based method using height, the length between the head and the pelvis, the maximum length between the pelvis and the feet, and the distance between both feet.

That is, after obtaining a binary silhouette video sequence from a video of a person walking with a vision camera, obtaining a normalized silhouette image from the silhouette video sequence, extracting identifiable gait feature information from the normalized silhouette image, and then extracting feature information. Objects can be identified by comparing them with the characteristic information stored in the computer.

Continuously, by using the image information obtained by filming by a plurality of CCTV cameras (C2) installed in the arrival hall without overlapping, the facial features, color and texture characteristics, and gait characteristics of the entrants located in the arrival hall are extracted to provide object feature information in real time. It is created and stored as (S107).

Next, the identity and similarity between the generated object characteristic information of the criminal traveler and the object characteristic information of the entrants in the arrival hall is detected (S109).

Here, the identity and similarity of facial features can be detected by feature point matching, which finds similar feature points by comparing extracted facial feature images of high-risk travelers with feature point descriptors extracted from facial feature images of immigrants in the arrival hall.

At this time, the identity and similarity may be determined to be the same or similar when, as a result of cross-checking of the object feature information, the RGB information matches more than a predetermined threshold such as 70% or more matching (S111).

In addition, the identity and similarity of color features are determined by generating a color histogram from the extracted color feature images of high-income travelers and color feature images of immigrants in the arrival hall, segmenting and extracting the main image features, comparing them to each other, and reaching a preset threshold. It can be detected only in the case of abnormalities.

In addition, the identity and similarity of the texture features are determined by image segmentation of the extracted texture feature images of high-income travelers and texture feature images of immigrants in the arrival hall, constructing MBR (Minimum Boundary Rectangle) of the object, and within each image. Texture features can be generated and detected by applying wavelet transform to the included object's MBR.

In addition, the identity and similarity of the gait features are obtained by obtaining a binary silhouette video sequence from image information obtained by shooting by a plurality of CCTV cameras (C1) installed at the immigration checkpoint, and then obtaining a normalized silhouette image from the silhouette video sequence , After extracting identifiable gait feature information from the normalized silhouette image, the extracted gait feature information is compared with the gait feature information of the image information obtained by shooting a plurality of CCTV cameras (C2) installed in the arrival hall and detected. can do.

Subsequently, an entrant whose similarity in at least one of the object characteristic information of the entrant in the arrival hall is equal to or higher than a predetermined threshold is recognized and identified as the pertinent bad traveler located in the arrival hall (S113).

In other words, through mutual comparison and analysis between facial features, color and texture characteristics, and gait characteristics of the high-risk traveler and the entrant in the arrival hall, the degree of similarity between the high-risk traveler detected in advance and the entrant in the arrival hall to be compared is analyzed. judge the degree of similarity.

For example, if a pre-detected criminal traveler disguises himself or changes clothes, such as wearing heavy makeup, glasses, or a mask, in the arrival hall, the gait characteristics of those entering the arrival hall differ from those of the pre-detected criminal travelers, a predetermined threshold of 70%. A person similar to the above can be judged as a bad traveler.

Thereafter, object feature information and image information of the specified criminal traveler are additionally stored, and the corresponding image information is captured and provided in real time to the customs situation room, customs rover, and marshal through a wireless communication network (S115). ).

Next, the self-driving robots (R) in the arrival hall extract the facial features, color and texture features, and gait features of the entrants from the image information obtained by filming with cameras to create and store object feature information in real time (S117).

In addition, identity and similarity between the object characteristic information of the entrants in the created arrival hall and the object characteristic information of the already specified high-risk traveler are detected (S119).

Here, the identity and similarity of facial features can be detected by feature point matching, which finds similar feature points by comparing extracted facial feature images of high-risk travelers with feature point descriptors extracted from facial feature images of immigrants in the arrival hall.

At this time, the identity and similarity may be determined to be the same or similar when the object feature information is matched at a predetermined threshold or higher, such as a match of 70% or more of the RGB information as a result of mutual comparison (S121).

In addition, the identity and similarity of color features are determined by generating a color histogram from the extracted color feature images of high-income travelers and color feature images of immigrants in the arrival hall, segmenting and extracting the main image features, comparing them to each other, and reaching a preset threshold. It can be detected only in the case of abnormalities.

In addition, the identity and similarity of the texture features are determined by image segmentation of the extracted texture feature images of high-income travelers and texture feature images of immigrants in the arrival hall, constructing MBR (Minimum Boundary Rectangle) of the object, and within each image. Texture features can be generated and detected by applying wavelet transform to the included object's MBR.

In addition, the identity and similarity of the gait features are obtained by obtaining a binary silhouette video sequence from image information obtained by shooting by a plurality of CCTV cameras (C1) installed at the immigration checkpoint, and then obtaining a normalized silhouette image from the silhouette video sequence , After extracting identifiable gait feature information from the normalized silhouette image, the extracted gait feature information is detected by comparing the extracted gait feature information with the gait feature information of the image information obtained by shooting a plurality of autonomous robots (R) disposed in the arrival hall. can do.

Thereafter, an entrant detected as matching at least one of the object characteristic information of the entrant in the arrival hall to a predetermined threshold or higher is recognized and re-specified as a bad traveler (S123).

In other words, through mutual comparison and analysis between facial features, color and texture characteristics, and gait characteristics of already specified high-risk travelers and entrants in the arrival hall, the degree of similarity between the high-risk travelers detected in advance and the entrants in the arrival hall to be compared is determined. is analyzed to determine the degree of similarity.

For example, if a pre-detected criminal traveler disguises himself or changes clothes, such as wearing heavy makeup, glasses, or a mask, in the arrival hall, the gait characteristics of those entering the arrival hall differ from those of the pre-detected criminal travelers, a predetermined threshold of 70%. A person similar to the above can be judged as a bad traveler.

In this process, the location of the re-specified criminal traveler is confirmed using the image information received from the autonomous robots (R), and relative location information corresponding to the distance between the re-specified criminal traveler and a specific place in the arrival hall; That is, based on the relative location information corresponding to the distance between a specific place such as a duty-free shop, toilet, store, quarantine facility, sculpture, aisle, etc. in the arrival hall and a high-risk traveler, if the relative location information is out of the standard range, the self-driving robot (R) and the moving direction of another autonomous driving robot (R) closest to the current location of the re-specified prime traveler or a specific place such as a duty-free shop, restroom, store, quarantine facility, sculpture, passage, etc. in the arrival hall closest to it ( or destination), and the location of the re-specified criminal traveler can be reconfirmed using the image information received from the other self-driving robot R.

For example, if the location of a re-specified criminal traveler is confirmed in the arrivals hall and the self-driving robot (R) being tracked goes out of the assigned (allocated) area or predetermined route and chases or follows closely, the likely tourist is not noticed. Since it can be grabbed and escaped or evaded, command and control to hand over to the nearest other self-driving robot (R), and at the same time, the other self-driving robot (R) is to confirm the location of the re-specified trespasser traveler and to move and track it. command and control.

That is, if the re-specified offender traveler falls more than 5M away from the predetermined sculpture A, it is determined that the self-driving robot R is out of the assigned (allocated) area, set route, or service area, and the re-specified offender traveler is re-specified. Control to move other self-driving robots (R) that have been assigned (assigned) to the current location or the service area that includes the sculpture A in advance, or are located closest to the current sculpture A, toward the re-specified prime traveler or toward the sculpture A. By doing so, it is possible to efficiently respond to the variable on-site situation, such as continuously identifying and flexibly tracking the re-specified high-crime traveler or the re-specified high-crime traveler without any suspicion.

Then, based on the image information received from the self-driving robots R and GPS, each location is recognized in real time, and the location information is displayed on the display unit 28 in real time (S125).

In addition, the image information obtained by photographing the criminal traveler in the re-specified arrival hall is displayed on the display unit 28 in real time.

At this time, the position of the bad tourist can be checked in real time on the electronic map using the unique ID or identifier assigned to the self-driving robot R that provides image information including the re-specified bad tourist, and the display unit 28 ) as an alarm popup.

Furthermore, object feature information and image information of re-specified high-risk travelers are additionally stored, and the image information is captured and provided in real time to the customs situation room, customs rover, and marshal through wireless communication networks and terminals. A message associated with the object characteristic information of the re-specified criminal traveler may be transmitted.

Therefore, even if a high-risk traveler who has passed through the immigration checkpoint at an airport or port disguises himself in the arrival hall by wearing heavy makeup, glasses, or a mask to avoid CIQ customs clearance, or changes clothes, he is closely observed by responding and tracking more efficiently along the route of his movement. and continuously monitored.

In addition, the self-driving robot (R) collects and analyzes the object feature information of the entrant using the image information obtained by filming with the camera while moving through the target area in the arrival hall. Even if a high-risk traveler hides or is located in a high-risk traveler, the location and movement of the high-risk traveler can be checked in real time, thereby improving the reliability and tracking efficiency of the high-risk traveler detection.

Meanwhile, the present invention can be implemented as a program that is executed by one or more processes in a computer and can be stored in a computer-readable medium (recording device for storing data).

That is, the present invention is implemented as a computer program and can be recorded on a computer-readable recording medium, and a program for implementing the operation by the method is recorded and the computer-readable recording medium can be read by a computer. It may include all types of recording devices in which data is stored.

On the other hand, the present invention is not limited by the above-described embodiments (embodiment) and the accompanying drawings, and can be variously modified and applied in various ways not illustrated within the scope without departing from the technical spirit of the present invention, as well as each It is clear to those of ordinary skill in the art that the components can be widely applied by substitution of components and changes to other equivalent embodiments.

Therefore, contents related to the modification and application of the technical features of the present invention should be interpreted as being included within the technical spirit and scope of the present invention.

11: Criminal traveler information input unit 12: First image information input unit
13: criminal traveler detection unit 14: first feature extraction unit
15: second image information input unit 16: second feature extraction unit
17: 1st similarity detection unit 18: 1st criminal traveler specific unit
19: storage unit 20: transmission unit
21: control unit 22: data communication unit
23: third feature extraction unit 24: second similarity detection unit
25: second criminal traveler specific unit 26: location recognition unit
27: GIS display unit 28: display unit
29: event alarm

Claims (4)

a bad traveler information input unit 11 that receives bad traveler information among entrants by comparing prior analysis information on the entrant with bad traveler database of the traveler information system;
A first image information input unit 12 receiving image information from a plurality of CCTV cameras C1 installed at the immigration checkpoint;
A bad traveler detection unit that identifies an entrant's object from the image information input through the first image information input unit 12 and compares it with the bad tourist information input through the bad traveler information input unit 11 to detect a bad traveler. (13);
a first object feature information extraction unit 14 that generates and stores object feature information in real time by extracting facial features, color and texture features, and gait features of the detected criminal traveler;
A second image information input unit 15 receiving image information from a plurality of CCTV cameras C2 installed in the arrival hall;
A second object feature information extraction unit (16) for generating and storing object feature information in real time by extracting facial features, color and texture features, and gait features of the immigrants from the image information input through the second image information input unit (15). );
A device for detecting the degree of identity and similarity between the object characteristic information of a frequent traveler generated by the first object characteristic information extraction unit 14 and the object characteristic information of entrants in the arrival hall generated by the second object characteristic information extraction unit 16 1 similarity detection unit 17;
a first bad traveler specific unit 18 that recognizes the entrants whose similarity value detected by the first similarity detection unit 17 is higher than a predetermined threshold value among the entrants in the arrival hall and identifies them as the pertinent bad traveler;
Identification information and arrangement information for each of the plurality of CCTV cameras (C1) installed at the immigration checkpoint, the plurality of CCTV cameras (C2) installed in the arrival hall, and the plurality of self-driving robots (R) disposed in the arrival hall, a storage unit (19) for storing object characteristic information and image information of a specified crime traveler;
a transmission unit 20 that provides object characteristic information and image information of a specific high-risk traveler to a customs control room, a customs rover, and a marshal;
a control unit 21 that controls functions of the self-driving robots R and overall operation of the system;
a data communication unit 22 that wirelessly communicates data with each of the autonomous robots R;
From the image information received through the data communication unit 22 and photographed by the cameras of the self-driving robots R, facial features, color and texture features, and gait features of entrants are extracted to generate and store object feature information in real time. a third object feature information extraction unit 23;
A second step for detecting the degree of identity and similarity between the object feature information of the prime traveler specified by the first bad traveler specification unit 18 and the object feature information of the entrants in the arrival hall generated by the third object feature information extraction unit 23 a similarity detection unit 24;
a second bad traveler specification unit 25 for recognizing an entrant whose similarity value detected by the second similarity detecting unit 24 is equal to or higher than a predetermined threshold value and re-specifying the entrant as a bad traveler;
a location recognition unit 26 that recognizes each location in real time based on the video information received from the self-driving robots and GPS;
a GIS display unit 27 for displaying location information of the self-driving robots R on the display unit 28; and
an event alarm unit 29 that outputs image information including a criminal traveler obtained by photographing with the cameras of the self-driving robots R to the display unit 28 in real time;
A system for tracking criminal travelers in the arrivals hall, including
According to claim 1,
The control unit 21 receives image information from the self-driving robots R, and uses the received image information to confirm the location of the re-specified criminal traveler,
Based on the relative location information corresponding to the distance between the re-specified high-risk traveler and a specific place in the arrival hall, if the relative location information is out of the standard range, the self-driving robot R is the closest to another autonomous robot. (R) is controlled to change the moving direction to the specific place, and the location of the re-specified criminal traveler is reconfirmed using the image information received from the other self-driving robot (R). tracking system.
How to track a high-risk traveler in the arrivals hall, including each of the following steps.
(a) Searching and designating bad travelers among the entrants by comparing preliminary analysis information on entrants with bad traveler database of the tourist information system.
(b) Recognizing the face of an entrant passing through immigration using image information obtained from a plurality of CCTV cameras installed at the immigration checkpoint, and identifying and detecting a bad traveler by comparing it with designated bad traveler information
(c) generating and storing object feature information in real time by extracting facial features, color and texture features, and gait features of the detected criminal traveler;
(d) Extracting facial features, color and texture features, and gait features of entrants from image information obtained by filming by a plurality of CCTV cameras installed in the arrival hall to generate and store object feature information in real time
(e) Detecting the degree of identity and similarity between the generated object characteristic information of the criminal traveler and the object characteristic information of the entrants in the arrival hall
(f) Recognizing an entrant whose similarity in at least one of the object characteristic information of the entrant in the arrival hall is detected as higher than a predetermined threshold and specifying the entrant as a pertinent bad traveler
(g) additionally storing the object characteristic information and image information of the specified criminal traveler and providing them to the customs control room, customs rover, and marshal
(h) Step of generating and storing object feature information in real time by extracting facial features, color and texture features, and gait features of entrants from image information captured by self-driving robots in the arrival hall using cameras
(i) Detecting the degree of identity and similarity between the object characteristic information of the entrants in the created arrival hall and the object characteristic information of the already specified high-risk traveler
(j) Recognizing an entrant whose similarity in at least one of object characteristic information of the entrant in the arrival hall is detected as higher than a predetermined threshold and re-specifying the entrant as a high-risk traveler
(k) Recognizing each position in real time based on video information received from autonomous robots and GPS, and displaying the position information on a display unit
According to claim 3,
In the step (j), the location of the re-specified criminal traveler is confirmed using the image information received from the self-driving robots;
Based on the relative location information corresponding to the distance between the re-specified criminal traveler and a specific place in the arrivals hall, if the relative location information is out of the standard range, the moving direction of the self-driving robot and the other autonomous robot closest to it and re-confirming the location of the re-specified criminal traveler using image information received from the other self-driving robot.

Images