KR102272850B1 - multi table type cognitive training device using artificial intelligence and method for player cognition using the same - Google Patents

Abstract

Disclosed are a multi-table type cognitive training device, and a player recognition method using the same. According to the present invention, the disclosed multi-table type cognitive training device comprises a display unit, a base unit, and a management system.

Description

Multi table type cognitive training device using artificial intelligence and method for player cognition using the same

The present invention relates to a multi-table type cognitive training apparatus using artificial intelligence, and more particularly, artificial intelligence that allows a plurality of users to be provided with cognitive training and cognitive stimulation as a content game at the same time. It relates to a multi-table type cognitive training apparatus using the same and a player recognition method using the same.

As the life expectancy of people is significantly extended, the incidence of geriatric diseases is also rapidly increasing. According to US statistics, the death rate from major diseases such as cancer and cardiovascular disease has been steadily decreasing due to improved treatment methods, but it is reported that the death rate from dementia is continuously increasing.

In fact, according to the Alzheimer's Association of America, the death rate from dementia has increased by 89% since 2000. With the aging population, the incidence of dementia is also increasing, and the cause of dementia is still unknown. However, the information related to dementia known so far is as follows.

When analyzing the brains of dementia patients, a representative example was the fact that agglomerated masses (plaques) of amyloid beta protein were observed as a kind of waste product along the periphery of brain cells. Based on this, studies to remove amyloid beta have been actively conducted at one time, but it has not been established as an alternative to solving dementia. This is because, in many cases, people's cognitive ability did not improve even if amyloid beta was removed, and even those with many plaques (clumps) did not have dementia.

In addition, it has been reported that people with early dementia are more likely to recover from dementia through continuous cognitive training and cognitive stimulation rather than treatment with the aforementioned drugs.

For this reason, many complex stimulation devices and methods for preventing dementia by providing physical activity and brain stimulation at the same time have been proposed, and Korean Patent Laid-Open No. 2009-0067578 is also one of these technologies. In this technology, game content such as language, math, common sense, etc. is provided to the user through a monitor, and when the user takes a specific reaction action according to the instruction of the provided content, the result is fed back to the user.

The advantage of this technology is that it can increase the concentration of education in that it can provide users with various types of content games visually. However, it stores the results of repeated tests without user information and provides content for each level. way to provide

In addition, the technology of Korean Patent Laid-Open No. 2009-0067578 is limited to individual dementia patients because it is a method to provide brain stimulation to a specific user using game contents. In other words, there is a problem that people with similar dementia conditions cannot provide cognitive ability training or cognitive stimulation in a cooperative way.

Republic of Korea Patent Publication No. 2009-0067578

An object of the present invention is to provide a multi-table type cognitive training apparatus using artificial intelligence that can provide various cognitive training and cognitive stimulation such as visual, auditory and physical movement to a patient with dementia or a patient with low cognitive ability, and a player recognition method using the same. is in providing.

In addition, it is an object of the present invention to provide a multi-table type cognitive training apparatus using artificial intelligence that allows a plurality of dementia patients or patients with low cognitive ability to simultaneously receive cognitive training and cognitive stimulation in a cooperative manner, and a player recognition method using the same. is in providing.

In addition, an object of the present invention is to analyze various touch patterns input by players into a display area while playing a content game through a learning model based on deep learning, and then an artificial object that can accurately recognize an object or a player inputting a touch pattern. An object of the present invention is to provide a multi-table type cognitive training apparatus using intelligence and a player recognition method using the same.

In addition, an object of the present invention is to provide a multi-table type cognitive training apparatus using artificial intelligence capable of distinguishing between players who provide a touch input in a content game based on the motion of a player who plays the content game and patterns input to the display area, and An object of the present invention is to provide a player recognition method using the same.

The multi-table type cognitive training apparatus of the present invention includes: a display unit including a touch panel for displaying content games for providing cognitive training or cognitive stimulation to a user and a plate on which the touch panel is seated; and a base unit supporting the display unit and a management system controlling the display unit and the base unit.

In addition, the management system, the touch information is input and the display unit for displaying the content game; a storage unit for storing information on content games and information on users; a driving unit that moves the display unit up and down, rotates and tilts; a content providing unit providing a content game for cognitive training and a content game for cognitive stimulation training to a user; a recognition unit for recognizing a touch pattern based on the touch information input through the display unit; a determination unit which determines whether the touch pattern recognized by the recognition unit is touch information input by which user from among a plurality of users; a first sensor unit disposed along a periphery of a side surface of the display unit to obtain location information of the user; a second sensor unit disposed on the inside corresponding to the display area of the display unit to obtain face information of the user; a third sensor unit disposed along a non-display area between the touch panel and the plate of the display unit to obtain a position of a user's marker and information included in the marker; and a fourth sensor unit configured to obtain movement information of a plurality of users who play a content game in an upper space of the display unit.

In addition, the determination unit of the management system based on the touch information input through the display unit, the user's face image and motion image provided by the second sensor unit and the fourth sensor unit, a learning model built by machine learning extracting the user's face and motion images corresponding to the touch information using the image, and using the image extracted from the learning model and the user's location information provided from the first sensor unit to respond to the touch information input to the display unit. Characterized in determining the user.

In addition, the determination unit of the multi-table type cognitive training apparatus of the present invention, an image separation unit for separating the image captured by the second sensor unit and the fourth sensor unit into an image centered on the user's face and an image centered on the user's motion ; a facial image conversion unit for converting a face image among the images separated by the image separation unit; a motion image conversion unit for converting a motion image among the images separated by the image separation unit; a pattern image conversion unit for converting touch information input to the display unit; a learning model constructed by machine learning to extract a face image and a motion image corresponding to each user's touch information based on the images converted by the face image conversion unit, the motion image conversion unit, and the pattern image conversion unit; a separation extraction unit for separating and extracting a user, a user's motion, and a tool used by the user from the learning model; and a determining unit configured to determine a user corresponding to the touch information input to the display unit by using the image extracted from the separation extraction unit and the user's location information by the first sensor unit.

In addition, the determination unit of the multi-table type cognitive training apparatus of the present invention matches the image extracted from the separation extraction unit and the location information provided by the first sensor unit to check which user the touch information is input by. It further includes an inspection unit.

Here, the touch information input to the display unit may be information input to a touch point of a finger of a user who plays a content game or a marker attached to a tool used to play a content game.

In addition, the player recognition method using the multi-table type cognitive training apparatus of the present invention, a display unit having a touch panel for displaying content games and a plate on which the touch panel is seated; In a player recognition method using a multi-table type cognitive training apparatus comprising a base unit supporting the display unit and a management system for controlling the display unit and the base unit, the management system receives touch information and displays a content game a display unit, a storage unit storing information about content games and information about a user, a driving unit for moving the display unit up and down, rotation and tilt, and content games and cognitive stimulation training for cognitive ability training to the user A content providing unit providing a content game, a recognition unit recognizing a touch pattern based on touch information input through the display unit, and a touch input by any one of a plurality of users to which the touch pattern recognized by the recognizing unit is recognized A determination unit for determining whether information is information, a first sensor unit disposed along a periphery of the side surface of the display unit to obtain the user's location information, and an inner side corresponding to the display area of the display unit to provide information on the user's face a second sensor unit to obtain a, a third sensor unit disposed along a non-display area between the touch panel and the plate of the display unit to obtain a position of a user's marker and information included in the marker; The upper space includes a fourth sensor unit that acquires movement information of a plurality of users who play a content game, and the determination unit includes each user for a plurality of users provided from the second sensor unit and the fourth sensor unit. Separating a face image and a user's motion image; converting the separated user's face image and motion images into images that can be learned by a learning model built by machine learning; converting the touch information input to the display unit into an image that can be learned by a learning model built by machine learning; After performing machine learning on the converted user's face image, motion image, and touch information images through a learning model, extracting a user face image, a motion image, and a touch information image corresponding thereto for each user; and after examining each user corresponding to the touch information based on the extracted images and location information for each of the plurality of usage fields provided by the first sensor unit, each user corresponding to the touch information Characterized in performing the step of determining.

The multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention can provide various cognitive training and cognitive stimulation such as visual, auditory and physical movement to a patient with dementia or a patient with low cognitive ability there is

In addition, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention are effective in that a plurality of dementia patients or patients with low cognitive ability can simultaneously receive cognitive training and cognitive stimulation in a cooperative manner. there is

In addition, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention analyze various touch patterns input to the display area by players while playing the content game through a learning model based on deep learning. , there is an effect of accurately recognizing an object or player inputting a touch pattern.

In addition, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention provide a player providing a touch input in a content game based on the motion of a player who plays the content game and patterns input to the display area have the effect of distinguishing them.

1 is a diagram showing how users use the multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention.
2 is a perspective view showing the structure of a multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention.
3 is a block diagram of a multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention.
4A to 4D are diagrams illustrating a state in which a multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention provides a play area according to the number of users.
5 is a diagram for explaining a process of recognizing a user or an object based on a touch input input to a display unit by the multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention.
6 to 8 are diagrams for explaining the principle of recognizing a user using a multi-table type cognitive training apparatus using artificial intelligence and an object used in a provided content game according to an embodiment of the present invention.
9 is a diagram for explaining a user's touch operation and an object operation of content in the multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention.
10 is a diagram illustrating a state in which a multi-table type cognitive training apparatus using artificial intelligence recognizes a touch position of a display area according to an embodiment of the present invention.
11 is a diagram illustrating a state in which a possible touch pattern CP is estimated from the touch contact units TC input to the display area DA.
12 is a diagram illustrating a touch ID corresponding to touch information such as the touch contact unit TC.
13 is a diagram in which a function corresponding to a touch ID is applied to a content game played in the display area DA.
14 is a block diagram illustrating a detailed configuration of a recognition unit of a multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention.
15 is a detailed block diagram of a determination unit of a multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention.
16 is a view for explaining a state in which the multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention recognizes a plurality of players.

Since the present invention can apply various transformations and can have various embodiments, specific embodiments are illustrated in the drawings and described in detail in the detailed description. Effects and features of the present invention, and a method of achieving them, will become apparent with reference to the embodiments described below in detail in conjunction with the drawings. However, the present invention is not limited to the embodiments disclosed below and may be implemented in various forms. In the following embodiments, terms such as first, second, etc. are used for the purpose of distinguishing one component from another, not in a limiting sense. Also, the singular expression includes the plural expression unless the context clearly dictates otherwise. Also, terms such as include or have means that the features or components described in the specification are present, and do not preclude the possibility that one or more other features or components will be added. In addition, in the drawings, the size of the components may be exaggerated or reduced for convenience of description. For example, since the size and thickness of each component shown in the drawings are arbitrarily indicated for convenience of description, the present invention is not necessarily limited to the illustrated bar.

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

1 is a diagram showing how users use the multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention. 2 is a perspective view showing the structure of a multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention.

1 and 2, the multi-table type cognitive training apparatus 100 using artificial intelligence of the present invention includes a display unit 100a and a base unit 100b supporting the display unit 100a. The display unit 100a includes a display area DA that provides various content games to a user and a non-display area NDA disposed along the periphery of the display area DA. A display panel capable of displaying a content game may be disposed in an area corresponding to the display area DA. The display panel may be a touch screen panel (TSP) implemented with a liquid crystal display (LCD) or an organic light emitting display (OLED). In addition, the multi-table type cognitive training apparatus 100 of the present invention is a personal computer, laptop, notebook, netbook or tablet, personal digital assistant (PDA), navigation, home appliance (home appliance), image processing It may be implemented in a device, such as a device, a smartphone, a 3D TV, or a digital information display (DID). In particular, each of the components of the present invention may be operated within a processor and a memory disposed within the devices described above.

Recognition means for recognizing a user may be disposed in the non-display area NDA. For example, a plurality of authentication pads N1 , N2 , N3 , and N4 for checking information of a user who uses a content game may be disposed. The authentication pads N1, N2, N3, and N4 may function to recognize a card including a user's unique identifier information and provide a customized content game based on the recognized user's information. In addition, when the first to fourth authentication pads N1, N2, N3, and N4 recognize different information for a plurality of users, based on this, a content game that a plurality of users can play together can be recommended or provided. have. That is, the multi-table type cognitive training apparatus 100 of the present invention provides a cognitive training content game or a cognitive stimulation content game required for each of a plurality of users, or recommends and provides a content game that a plurality of users can play together. can do. In the present specification, a user using the multi-table type cognitive training apparatus 100, a player using a content game, and a user using a service provided by the multi-table type cognitive training apparatus 100 are used without distinction. Accordingly, a player who plays the content game may be named as a user or a user because the player who uses or uses the multi-table type cognitive training apparatus 100 .

In addition, the multi-table type cognitive training apparatus 100 using artificial intelligence of the present invention provides a content game performed by a plurality of users (P1, P2, P3, P4) together or each content game suitable for individual users. , the display area DA may be divided into first to fourth game areas 110a, 110b, 110c, and 110d.

A plurality of users (players: P1, P2, P3, P4) may place a marker M in their divided first to fourth game areas 110a, 110b, 110c, and 110d, respectively, and play a content game. Also, the display area DA may include a common display area CDA. The common display area CDA is commonly applied to all of the plurality of users (players: P1, P2, P3, P4) when a plurality of users (players: P1, P2, P3, P4) play one content game together. It may be a game area. Also, when a plurality of users (players: P1, P2, P3, P4) play an individual content game, it may be an area commonly used by individual content games.

Referring to FIG. 2 , the display unit 100a of the multi-table type cognitive training apparatus 100 using artificial intelligence of the present invention may be implemented in the form of a rectangular plate or a table. The upper surface of the display unit 100a includes a display area DA and a non-display area NDA for displaying content games. In addition, the first sensor unit 130a in which a plurality of sensors are arranged at regular intervals may be disposed on four sides of the display unit 100a. The first sensor unit 130a functions to obtain location information of a user for playing a content game. Accordingly, the sensors constituting the first sensor unit 130a may be any one or a mixture of at least one of an ultrasonic sensor, an infrared sensor, a LIDAR sensor, a radar sensor, and a camera sensor.

A second sensor unit 130b capable of acquiring face information (facial information) of users (players) may be disposed in the display area DA of the display unit 100a. In the drawing, four sensors are disposed in areas corresponding to each corner of the display unit 100a having a rectangular structure, but this may be changed according to the number of a plurality of users playing a content game. For example, when six players play a content game, the number of sensors constituting the second sensor unit 130b may be six or more.

The third sensor unit 130c may be disposed on an upper surface corresponding to each corner of the non-display area NDA of the display unit 100a. When the ID card including the user's unique identifier information is placed, the third sensor unit 130c obtains information on the ID card of the user (player) to provide a content game or a play area in which to play the content game according to the number of ID cards can be used to partition Also, the third sensor unit 130c may serve to recognize the presence or absence of a player who is currently playing a content game provided based on a state in which the player's ID card is placed. In particular, the area of the third sensor unit 130c may partially overlap the display area DA.

In the non-display area NDA of the display unit 100a, first to fourth authentication pads N1, N2, N3, and N4 may be disposed. Although they are respectively arranged in the four corner areas of the display unit 100a in the drawing, this is not fixed. In particular, even if the user marker M sequentially contacts the authentication pad by the control unit connected to the first to fourth authentication pads N1, N2, N3, and N4, information including unique identifier information for each user is drawn. If possible, only one authentication pad (N) may be disposed.

In addition, in the multi-table type cognitive training apparatus 100 using artificial intelligence of the present invention, the fourth sensor unit 130d may be disposed in the space above the display unit 100a. The fourth sensor unit 130d may continuously acquire subject information of players who play the content game and reflect it in the content game, or obtain movement information of a player (user) and use it as information for determining the player's cognitive ability. A sensor constituting the fourth sensor unit 130d may be formed of a charge coupled device (CCD) or a complementary metal semiconductor (CMOS).

The base unit 100b includes a first support 120 that supports the display unit 100a and can vertically move, rotate, and tilt the display unit 100a, and is disposed below the first support 120 . and a second support 121 in contact with the . Components for controlling the operation of the multi-table type cognitive training apparatus 100 to be described later may be disposed inside the first support unit 120 . In addition, the second support part 121 may include a plurality of bridge parts so as to distribute the loads of the display unit 100a and the first support part 120 while being coupled to the first support part 120 . As shown in the drawing, the moving means 123 may be disposed on each bridge unit.

In addition, although not specifically illustrated in FIGS. 1 and 2 , the display unit 100a may be disposed in a horizontal direction or may be disposed in a vertical inclination direction by the first support unit 120 . Accordingly, the first support unit 120 may be coupled to the rear surface of the display unit 100a and a shaft capable of moving in the vertical direction may be disposed. The shaft may perform an up/down operation in the vertical direction by the power provided by the driving unit, and the display unit 100a is placed on the ground in response to the up/down operation of the shaft. It may be erected in a vertical direction or inclined to have a predetermined angle with respect to the ground.

As such, when the display unit 100a is erected or inclined, it can be used in a content game in which players are positioned on both sides with the display unit 100a interposed therebetween.

In addition, the display unit 100a may be provided with an event providing unit capable of providing a vivid sense of presence to players in the process of playing a content game. The event providing unit may be composed of a plurality of air nozzles or vibration motion or an intelligent speaker.

As such, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention can provide various cognitive training and cognitive stimulation such as visual, auditory and physical movements to dementia patients or patients with low cognitive ability. can have an effect.

In addition, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention are effective in that a plurality of dementia patients or patients with low cognitive ability can simultaneously receive cognitive training and cognitive stimulation in a cooperative manner. there is

In addition, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention analyze various touch patterns input to the display area by players while playing the content game through a learning model based on deep learning. , there is an effect of accurately recognizing an object or player inputting a touch pattern.

In addition, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention provide a player providing a touch input in a content game based on the motion of a player who plays the content game and patterns input to the display area have the effect of distinguishing them.

3 is a block diagram of a multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention. 4A to 4D are diagrams illustrating a state in which a multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention provides a play area according to the number of users. 5 is a view for explaining a process of recognizing a user or an object based on a touch input input to the display unit by the multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention. 6 to 8 are diagrams for explaining a principle of recognizing a user using a multi-table type cognitive training apparatus using artificial intelligence and an object used in a provided content game according to an embodiment of the present invention. 9 is a diagram for explaining a user's touch operation and an object operation of content in the multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention. 10 is a diagram illustrating a state in which a multi-table type cognitive training apparatus using artificial intelligence recognizes a touch position of a display area according to an embodiment of the present invention. 11 is a diagram illustrating a state in which a possible touch pattern CP is estimated from the touch contact units TC input to the display area DA, and FIG. 12 is touch information such as the touch contact unit TC. It is a diagram showing a touch ID corresponding to , and FIG. 13 is a diagram in which a function corresponding to the touch ID is applied to a content game played in the display area DA.

Referring to FIGS. 3 to 13 together with FIG. 2 , the multi-table type cognitive training apparatus 100 using artificial intelligence of the present invention includes a display unit 100a and a base unit 100b.

The multi-table type cognitive training apparatus 100 of the present invention may recognize users who play in order to provide a customized content game to the user or recognize tools used in the content game. In addition, the multi-table type cognitive training apparatus 100 of the present invention may provide customized cognitive training or a cognitive stimulation content game, respectively, according to the degree of cognitive ability for a plurality of users. The multi-table type cognitive training apparatus 100 includes first to fourth sensor units 130a, 130b, 130c, 130d, an input unit 301, a storage unit 302, a driving unit 303, a content providing unit 304, It may be operated by a management system including a control unit 305 , a recognition unit 306 , a determination unit 307 , and a display unit 308 .

As described above, the first sensor unit 130a acquires distance and location information between the user and the multi-table type cognitive training apparatus 100 for cognitive ability training, and the second sensor unit 130b allows the user to play a content game. Acquires facial (face) information before and after playing in a state positioned at one's place to play, and the third sensor unit 130c is a marker including an ID card containing user information or a card containing object information. The position of (M) and the information of the marker (M) are obtained, and the fourth sensor unit 130d obtains movement information of players (users) who play the content game.

In particular, the second sensor unit 130b and the fourth sensor unit 130d may include a camera including a color image sensor and a depth image sensor. In some embodiments, the depth image sensor may provide a depth image in which depth information of a scene is expressed as, for example, a 16-bit integer in units of [mm]. The number of bits for expressing depth information is not limited to 16 bits and may be changed, and may be provided in a real number type without being limited to an integer type. The depth image sensor may measure the distance from the camera to the object and the background by using infrared rays or the like to provide a depth image having a value proportional to or inversely proportional to the distance.

The pixel value of the depth image may be, for example, an integer of [mm] unit (but not limited thereto), instead of RGB color information.

The color image sensor is a device for acquiring a color image, and may be, for example, a complementary metal-oxide-semiconductor (CMOS) image sensor, a charge-coupled device (CCD) image sensor, or a stereo camera. The depth sensor may be a kinect, a time-of-flight (TOF) depth camera, or an optical 3D scanner, or the like.

The input unit 301 may include a touch screen panel because touch information may be input through the display unit 308 . Accordingly, the input unit 301 receives the information provided through the first to fourth sensor units 130a, 130b, 130c, and 130d and the display unit 308, prior information about the user, and information related to content games (pattern information). may be provided. Also, the input unit 301 may be implemented integrally with the display unit 308 .

The storage unit 302 may include information about content games and information about users. For example, when the user performs a test to determine the degree of dementia in advance, result information thereof may be stored. Information on content games may include information on characters played in each game and information on operations that can be performed in each game. The storage unit 302 may include a hard disk drive (HDD), a solid state disk (SSD), a silicon disk drive (SDD), a ROM, a RAM, a CD-ROM, a magnetic tape, a floppy disk, or an optical data storage device. . Storage 302 also includes instructions (eg, executable instructions), such as computer readable instructions or processor readable instructions. The instructions may include one or more instructions executable by a computer, such as by each of one or more processors.

For example, the one or more instructions may be executable by the one or more processors to perform operations comprising: determining a motion vector of the depth image and processing the depth image to encode the depth image. have.

As described above, the driving unit 303 physically moves the display unit 100a constituting the multi-table type cognitive training apparatus 100 or the first to fourth sensor units 130a, 130b, 130c, and 130d. It is possible to provide power to move the constituting sensors.

The content providing unit 304 may provide content games prepared for cognitive ability training and content games prepared for cognitive stimulation.

The recognizer 306 may recognize and analyze touch patterns based on the touch information provided from the display 308 , and then recognize an object corresponding to the analyzed pattern. Also, the recognizer 306 may recognize touch information provided by the player as well as touch information of a marker disposed on an object.

The determination unit 307 may obtain patterns and object information corresponding to the patterns recognized through the display unit 308 or filter the recognized patterns but not related to the content game currently in progress. In this way, by combining the selected object recognition information and the touch information provided from the player through the display unit 308, it is possible to finally determine which command is for which object.

Also, the recognition unit 306 and the determination unit 307 may determine the content games to be provided based on the number of users who will play the content game and information on the users. For example, when a plurality of users are authenticated by the multi-table type cognitive training apparatus 100, an individually customized content game may be provided based on the information of each user, or a content game that a plurality of users may progress together may be provided. .

Also, the recognition unit 306 and the determination unit 307 may divide the display area DA into a plurality of areas to correspond to a plurality of authenticated users, and then provide a game area to each user. At this time, the information input through the first to fourth sensor units 130a, 130b, 130c, 130d is synthesized without dividing the display area DA only when the user's marker is simply authenticated through the authentication pads. Split.

In particular, the multi-table type cognitive training apparatus 100 of the present invention includes a second sensor unit 130b and a fourth sensor unit when a plurality of users need to move along the circumference of the display unit 100a when a content game is played. Using 130d, a plurality of users (players) may be individually recognized and their positions may be tracked. In this way, when a player is recognized and tracked based on the image provided by the second sensor unit 130b and the fourth sensor unit 130d, when starting the first content game, the divided game area set for each player is displayed You can move to the position of the player along the edge of the area DA.

Accordingly, the player does not need to return to his/her initially set divided game area when there is a movement in the process of playing the content game. If the divided game areas set for a plurality of users (players) overlap each other, the game area is the closest area using the ultrasonic sensor information of the first sensor unit 130a and is the game area immediately adjacent to the player's game area. can be located. For example, if the second player is in the position of the fourth player in the process of playing the content game, the second player is placed in the adjacent area of the fourth player's game area so as not to overlap the fourth player's game area so that the next play can be performed do.

As described above, since the input unit 301 may be integrally formed with the display unit 308 , the control unit 305 , the storage unit 302 , the driving unit 303 , the content providing unit 304 , and the recognition unit 306 . , the determination unit 307 may be disposed in the base unit 100b of the multi-table type cognitive training apparatus 100 .

4A to 4D , the multi-table type cognitive training apparatus 100 using artificial intelligence of the present invention divides the display area DA into a plurality of divided game areas so as to correspond to the number of users using the content game. can do.

When there are two users to perform cognitive ability training or cognitive stimulation training through a content game, as shown in FIG. 4A , the display area DA is divided into an A1 game area and an A2 game area. Two users can play different content games in the divided game areas A1 and A2 or play one content game simultaneously in their own game areas.

As described above, in the case of a cooperative game or a competitive game in which one content game is played together, the user may move along the circumference of the display unit 100a while playing the game. In the multi-table type cognitive training apparatus 100 of the present invention, the divided game areas A1 and A2 may recognize and track the user to move the game area to a location corresponding to the current player's location. Accordingly, the player (user) playing the content game can proceed with the next play from the moving position without returning to the initially set divided game areas A1 and A2.

When there are three users to perform cognitive ability training or cognitive stimulation training through a content game, as shown in FIG. 4B , the display area DA is divided into an A1 area, an A2 game area, and an A3 game area. When there are four users to perform cognitive ability training or cognitive stimulation training through a content game, as shown in FIG. 4C , the display area DA is an A1 game area, A2 game area, A3 game area, and A4 game area. divided into regions. When there are five users to perform cognitive ability training or cognitive stimulation training through a content game, as shown in FIG. 4D , the display area DA is an A1 game area, A2 game area, A3 game area, and A4 game area. It is divided into area and A5 game area.

In addition, as described above, in the case of a cooperative game or a competitive game in which 3, 4, or 5 players play one content game together, when users move while playing the game, the divided game areas A1, A2, A3 , A4, A5) can move by tracking the direction of movement of the users. Accordingly, players (users) who play the content game can continuously play the next play from the play position without returning to the initially set or previously located game area.

The multi-table type cognitive training apparatus 100 using artificial intelligence of the present invention may recognize an object such as a tool used in a content game or a marker M that can identify a user (player).

5 is a diagram for explaining a method of recognizing a user's marker or a tool used in a game by the multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention. User markers or tool markers consist of physical objects. That is, it can be seen as recognition of an object rather than recognition of a part of the user's body.

Accordingly, the user marker M and the marker M attached to the tool include a touch point TP capable of providing touch information to the display unit 100a. The size of the touch point TP, the location of the touch point TP, and the number of touch points TP may be different for each marker M.

Referring to FIG. 5A , when three touch points TP are disposed on the marker M and each touch point TP is connected, a triangular touch pattern CP implemented by the touch points TP. can be obtained. Also, the area of each of the three touch points TP may be the same or at least one may be different. As shown in the drawing, the area occupied by the touch point TP is the touch area TA, the touch pattern CP formed by the connection of the touch points TP, and the area of the touch pattern CP is the pattern area. (PA), a direction TD of a touch point related to direction information for each touch point TP is defined.

The multi-table type cognitive training apparatus 100 using artificial intelligence of the present invention may obtain information provided by each marker M by analyzing the touch points TP disposed on the marker M. For example, figure information on whether the touch pattern CP is a triangle, a rectangle, or a pentagon may be acquired by using the touch point TP disposed on one marker M. That is, the multi-table type cognitive training apparatus 100 of the present invention may acquire information on the size and shape of a figure constituting the touch pattern CP. Also, information on the touch area TA for each touch point TP may be acquired. Also, with respect to the figure constituting the touch pattern CP, information on the direction of the touch point TP corresponding to each vertex may be obtained. For example, in the triangle of FIG. 5 (a), if virtual coordinate axes are formed in the center perpendicular to each other, the upper vertex is north (N), the lower left touch point (TP) is southwest (SW), and the lower right touch point is (TP) may obtain information about the direction (TD) of the touch point like the southeast (SE).

In this way, the multi-table type cognitive training apparatus using artificial intelligence of the present invention can distinguish each marker M based on various information provided by the touch points TP disposed on the marker.

In addition, the multi-table type cognitive training apparatus using artificial intelligence of the present invention is a touch area of the touch points (TP) disposed on the marker, the touch pattern (CP) formed by the touch points (TP), the touch pattern (CP) ), direction information of the touch point TP located at each vertex and a combination of these information can distinguish a plurality of markers from each other.

6 illustrates a principle of recognizing a marker M of a user using a content game provided by the multi-table type cognitive training apparatus 100 . Users who play a content game for cognitive ability training or a content game for cognitive stimulation are provided with markers M including different information so that they can be recognized as different players in the game. The marker M may be recognized by the authentication pads N1 , N2 , N3 , N4 of the above-described multi-table type cognitive training apparatus 100 or recognized by the marker recognition unit of the display unit 100a. As described above, each of the markers M provided to a plurality of users may be composed of touch points TP having different numbers and shapes, respectively.

As shown in FIGS. 6A to 6E , the same number of touch points TP are disposed in the first marker M1 and the second marker M2. However, the first marker M1 has a structure in which each of the touch points TP is disposed in a corner area adjacent to the touch point TP disposed at the edge of the first marker M1. , and the second marker M2 has a structure in which the touch points TP are disposed at both corners and the center of the second marker M2, respectively. Accordingly, the first marker M1 and the second marker M2 have different positions of the touch points TP, and the shape (figure shape) of the touch pattern CP based on the touch points TP disposed at each marker is also shown. It changes. Based on this information, the multi-table type cognitive training apparatus 100 of the present invention distinguishes the first marker M1 and the second marker M2.

Although only the first and second markers M1 and M2 on which the three touch points TP are disposed are illustrated in the drawing, a plurality of markers M including the three touch points TP may be formed. When there are a plurality of markers M composed of three touch points TP, it may be difficult to accurately distinguish only the position of the touch point TP and the shape of the touch panel CP. In this case, the touch area TA for each of the touch points TP disposed on each marker M is made different from each other, or any one or two of the three touch points TP is set to a different type of touch point TP. ) can be implemented. For example, one of the three touch points TP may be formed in a circle, and the other two may be formed in an ellipse or a triangle (or a rectangular structure), respectively.

Referring to FIGS. 6C and 6D , four touch points TP are disposed on the third marker M3 and the fourth marker M4 . The touch points TP of the third marker M3 are respectively disposed at four corners of the third marker M3 , and the touch points TP of the fourth marker M4 are located on one side of the fourth marker M4 . Two touch points TP are disposed at both corners, and two touch points TP are disposed between both edges of the other side facing one side. Accordingly, the touch pattern CP formed by the touch points TP of the third marker M3 has the same rectangular pattern as the shape of the third marker M3 and the touch point TP of the fourth marker M4. The touch pattern CP formed by the elements has a trapezoidal shape. That is, since the number of touch points TP disposed on the third marker M3 and the fourth marker M4 is the same, but the shape of the touch pattern CP is different, the two markers M3 and M4 are different markers. can be recognized. As described above, when there are a plurality of markers M composed of four touch points TP, the touch areas TA of the four touch points TP are formed differently, or the touch areas TA of the four touch points TP are formed differently. By combining the location coordinate information, the markers M may be recognized differently.

The fifth marker M5 of FIG. 6E has five touch points TP, and the touch pattern CP formed by these touch points TP has a pentagonal structure. Accordingly, the first to fourth markers M1 , M2 , M3 , and M4 of (a) to (d) may be distinguished from each other because the shape of the touch pattern CP is different. As described above, when there are a plurality of markers M composed of five touch points TP, the touch areas TA of the five touch points TP are formed differently, or the touch areas TA of the five touch points TP are formed differently. By combining the location coordinate information, the markers M may be recognized differently.

As shown in FIG. 7 , the multi-table type cognitive training apparatus 100 using artificial intelligence of the present invention recognizes which tools (horses) used in the chess game are when the provided content game is a chess game. can That is, a marker M may be attached to each piece of the chess game, and the number of touch points TP or a touch pattern CP composed of the touch points TP to distinguish each piece of the marker M. shape can be set differently.

In a chess game, pieces consisting of a king, queen, bishop, knight, rook, and pawn are placed on each side of 64 squares to take the opponent's king. Therefore, the multi-table type cognitive training apparatus 100 must be able to recognize the horses to proceed with the chess game by the player's play. In (a) to (d), the type of horse (tool) was recognized by attaching a marker (M) to the lower side of the king, queen, bishop and knight used in chess. Three touch points TP are arranged on the marker M attached to the king, and these form a triangular touch pattern CP. Four touch points TP are arranged on the marker M attached to the queen, and these are formed in the structure of a triangular touch pattern CP as shown in (a), but a touch point TP is additionally arranged in the center. It can be distinguished from the King's Marker (M). As described above, the recognition of the marker M includes the number of touch points TP, the touch pattern CP formed by the touch points TP, the area (shape) of each touch point TP, and the touch pattern. Since the areas of (CP) are recognized in combination, even if the number of touch points (TP) is the same, they can be recognized as different markers (M).

8 shows that the content game is a mole game, and when a player uses a hammer (tool), the players (Player1, Player2, Player3, Player4) can be distinguished by attaching markers (M) to both edge surfaces of the hammer. have.

Three touch points TP are arranged on the poof hammer marker M of the first player (player1), and 4 touch points TP are disposed on the poof hammer marker M of the second player (player2). ) is placed, 4 touchpoints TP are placed on the marker M of the bang hammer that the third player (player3) has, and 5 touch points TP are placed on the marker M of the bang hammer that the fourth player (player4) has. A touch point TP is disposed. As described in FIG. 7 , the number of touch points TP that each player has, the touch pattern CP formed by the touch points TP, the area (shape) of each touch point TP, and It can be recognized by combining information about the area of the touch pattern CP.

In particular, since the mole game as shown in FIG. 8 is progressed by each of the players holding their bang hammer and touching the display area DA where the moles are displayed, information about the touch point TP placed on the marker M and The movement information of the player obtained by the fourth sensor unit 130d disposed in the multi-table type cognitive training apparatus 100 may be used together.

That is, a touch point TP is disposed on the marker M disposed on the poof hammer, and when the marker M of the bang hammer comes into contact with the display area DA by the action of the player, the touch point TP is included. The display unit 308 acquires the marker (M) information. In this case, when a plurality of players (player1, player2, player3, player4) simultaneously use the bang hammer, movement information of the players may be additionally considered in order to obtain more accurate game result information.

For example, if there are balloons to be popped with a bang hammer or moles to be hit in the display area (DA), and they are separated to correspond to each player (player1, player2, player3, player4) (color, shape, etc.) , can accurately distinguish the bang hammer (player) that is in contact with (touched) balloons or moles, so that the game progress and game results can be shown. Therefore, when a plurality of infront hammers contact (touch) the balloon or mole in a specific area, the multi-table type cognitive training apparatus 100 of the present invention can accurately distinguish the matching between the infront hammer and the balloon or mole in the fourth The player's movement information by the sensor unit 130d is additionally considered. This will be described in more detail with reference to FIG. 15 .

FIG. 9 is a diagram for explaining touch information on a finger and a function in the content game for the touch information of a user (player) when using a user's (player) finger touch information when playing a content game. When the user (player) finger-touches (FT) a character or object of the content game with his index finger as in (a), the multi-table type cognitive training apparatus 100 of the present invention selects the touched character or object. It can be recognized as being and perform the set function.

When the first and second finger touches FT1 and FT2 are performed using the index and middle fingers as shown in (b), the multi-table type cognitive training apparatus 100 of the present invention is the size of the character or object touched in the content game. can be recognized and functioned as controlling When the first to third finger touches (FT1, FT2, FT3) are performed with the index finger, middle finger, and ring finger (or thumb, index finger and middle finger) as shown in (c), the multi-table type cognitive training apparatus 100 of the present invention is The touched character or object can be recognized as moving and functioned.

As shown in (d), when the first to fourth finger touches (FT1, FT2, FT3, FT4) are performed using the index finger, middle finger, ring finger, and little finger (or thumb, index finger, middle finger, and ring finger), the present invention The multi-table type cognitive training apparatus 100 may recognize and function as rotating a character or object touched in a content game.

(a) of FIG. 10 shows that the touch part touched on the display area DA is recognized in the form of position coordinates, and (b) of FIG. 10 shows that when the touch points TP are arranged like the marker M, the corresponding touch point ( For TP), ①, ②, and ③ are designated and recognized as one touch ID, and (c) is the area or shape of the touch point (TP) disposed on the marker (M) as described above. It represents the case of different cases and recognition of an area due to a finger touch (FT) by a human finger.

When a touch point TP or a touch by a finger is input to the display area DA, the multi-table type cognitive training apparatus 100 using the artificial intelligence of the present invention for the touch portion is the horizontal axis of the display area DA. (X coordinate) and the vertical axis (Y coordinate) to calculate the touch coordinates. Accordingly, when a touch input is provided to the display area DA, the multi-table type cognitive training apparatus 100 of the present invention recognizes the position coordinates of the touch part.

In addition, (b) shows that a finger touch FT by a finger or a plurality of markers M having the same number of touch points TP or different numbers of touch points TP in the display area DA is input. In this case, information such as a touch pattern (CP) may be acquired by numbering the corresponding touch contact unit as in ①, ②, or ③. That is, the touch ID may be set with respect to the touch pattern CP composed of the touch contact units based on the shape, size, and location information of each touch contact unit.

In addition, as in (c), based on the touch area TA recognized in the display area DA by the touch points TP disposed on the marker M and the finger touches FT by the finger, An object touched on the display area DA may be identified.

For example, when a marker M on which the touch points TP are disposed or a finger touch FT by a finger is input through the display area DA, the multi-table type cognitive training apparatus 100 of the present invention performs each touch The position coordinates (x, y) of the contact part are recognized. When a plurality of users play a content game, many touch contact units (such as a touch point or a touch by a finger) are recognized in the display area DA, and only valid touch information from among such touch information needs to be filtered. The marker information by the touch points TP of the marker M or the function information by the finger are preset and stored. Thereafter, when touch inputs are provided to the display area DA, it is necessary to extract only valid information from among a lot of touch information touched on the display area DA.

11 is a diagram illustrating a state in which a possible touch pattern CP is estimated from the touch contact units TC input to the display area DA, and FIG. 12 is touch information such as the touch contact unit TC. It is a diagram showing a touch ID corresponding to , and FIG. 13 is a diagram in which a function corresponding to the touch ID is applied to a content game played in the display area DA.

11 to 13 , in the display area DA, a tool used for a content game, a marker M of a player (user), and finger touch information of the player may be input. Accordingly, as shown in FIG. 11 , a large number of touch contact units TC can be recognized in the display area DA, and a plurality of touch patterns CP including the touch contact units TC are provided. may appear As shown in the drawing, when 16 touch contact units TC are recognized in the display area DA, touch patterns CP that can be recognized as 16 touch contact units TC are first to eleventh touch patterns CP. These are the touch patterns P1 to P11.

However, since the first to eleventh touch patterns P1 to P11 do not all provide information on the content game, the player's finger touch, or the player's marker M, it is necessary to select valid touch patterns among these touch patterns. have.

As shown in Figure 12, the multi-table type cognitive training apparatus 100 using artificial intelligence of the present invention is a touch pattern (CP) consisting of the touch contact unit (TC) shown in (a) to (d). It is saved in the form of touch ID (ID11, ID21, ID23, ID105). As described above, the touch ID information includes not only the shape of the touch pattern CP made of the touch contact units TC, but also information about the areas of each touch contact unit TC, the direction of the touch contact unit TC, The area information of the touch pattern CP is combined and stored as one touch ID information.

Accordingly, as shown in FIG. 11 , when a plurality of touch contact units TC are input to the display area DA, the first to eleventh touch patterns P1 to P11 including the touch contact units TC are recognized. After that, only information that is effective and appropriate for the currently provided content game is selected by comparing the recognized first to eleventh touch patterns P1 to P11 with pre-stored touch ID information.

As shown in FIG. 13 , among the first to eleventh touch patterns P1 to P11 recognized in FIG. 11 , a first touch pattern P1 : ID11 , a fourth touch pattern P4 : ID23 , and an eleventh touch pattern It can be seen that only (P11: ID21) and the seventh touch pattern (P7: ID105) are matched with information previously stored in the multi-table type cognitive training apparatus 100 . The seventh touch pattern P7 matches the touch pattern of the pre-stored information, but does not function on the display area DA when the touch pattern is not used in the current content game.

The first touch pattern (P1: ID11) performs a function of instructing a character (person) and moving in the direction in which it is located, and the fourth touch pattern (P4: ID23) is an object movement speed within the range in the content game. Performs an effect function of doubling, the eleventh touch pattern (P11: ID21) performs a function of instructing the vehicle and moving in the direction it is located, and the seventh touch pattern (P7: ID105) is currently executed There is no registered object or function data in the content game in progress.

Accordingly, as shown in FIG. 13 , among the first to eleventh touch patterns P1 to P11 displayed in the display area DA, the first touch patterns P1: ID11 and the fourth touch patterns P4: ID23 ) and the eleventh touch pattern (P11: ID21) can be seen operating in the content game.

As such, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention can provide various cognitive training and cognitive stimulation such as visual, auditory and physical movements to dementia patients or patients with low cognitive ability. can have an effect.

In addition, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention are effective in that a plurality of dementia patients or patients with low cognitive ability can simultaneously receive cognitive training and cognitive stimulation in a cooperative manner. there is

In addition, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention analyze various touch patterns input to the display area by players while playing the content game through a learning model based on deep learning. , there is an effect of accurately recognizing an object or player inputting a touch pattern.

In addition, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention provide a player providing a touch input in a content game based on the motion of a player who plays the content game and patterns input to the display area have the effect of distinguishing them.

14 is a block diagram illustrating a specific configuration of a recognition unit of a multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention. 15 is a detailed block diagram of a determination unit of a multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention. 16 is a diagram for explaining a state in which a multi-table type cognitive training apparatus using artificial intelligence according to an embodiment of the present invention recognizes a plurality of players.

14 and 15 together with FIGS. 2 and 3 , the multi-table type cognitive training apparatus 100 using artificial intelligence of the present invention includes the first to fourth sensor units 130a, 130b, 130c, and 130d. , an input unit 301 , a storage unit 302 , a driving unit 303 , a content providing unit 304 , a control unit 305 , a recognition unit 306 , a determining unit 307 , and a display unit 308 . The recognition unit 306 includes a pattern recognition unit 306a, a pattern analysis unit 306b, a pattern database 306c, for recognizing an object by recognizing and analyzing touch information input through the display unit 308, It includes an object recognition unit 306d and an object database 306e (Object Data Base).

As described above, the storage unit 302 disposed in the multi-table type cognitive training apparatus 100 of the present invention includes touch patterns CP of the markers M and touch patterns composed of a finger touch by the user's finger. (CP) information is stored. When touch information by a finger touch of a player (user) and touch points TP of a marker M is input through the display area DA of the display unit 308, the pattern recognition unit ( 306a).

As described in FIG. 11 , touch information is displayed in the form of a plurality of touch contact units TC in the display area DA by touch points TP disposed on the marker M and a touch provided by the player's finger. is input

The pattern analyzer 306b analyzes the touch patterns CP formed by the touch contact units TC based on the touch information determined as the effective touch area. That is, based on the pattern information stored in the pattern DB 306c, whether the touch patterns CP by the touch contact units TC input to the display area DA are a touch pattern indicating an object or a specific function (effect) Analyze whether it is a touch pattern that implements Although the pattern DB 306c is shown as a configuration of the recognition unit 306a, it may be implemented as a part of the storage unit 302 of the multi-table type cognitive training apparatus 100 .

After the pattern analyzer 306b analyzes the touch patterns CP of the touch contact units TC input to the display area DA, the results are provided to the object recognizer 306d. The object recognition unit 306d recognizes which object or specific function the touch patterns CP analyzed by the pattern analysis unit 306b are related to. The object recognition unit 306d compares the touch pattern CP provided by the pattern analysis unit 306b based on the object information (including specific effect and function information corresponding to the touch pattern) stored in the object database 306e. . That is, the object database 306e lists characters active in each content game, specific effect and function information that can be performed in the game, and touch patterns (CP) in the form of a lookup table. Accordingly, a character having the same touch pattern CP or a specific effect and function in a game may be matched with the touch patterns CP analyzed by the pattern analysis unit 306b. In this case, the object recognition unit 306d may receive the content information and recognize an object or a specific function currently used in the content game.

In addition, the object recognition unit 306d receives each content game information stored in the storage unit 302, and whether it is a touch pattern recognized as a touch pattern (CP), but is valid (used) in relation to the current content game. can be filtered. For example, if the pattern analysis unit 306b analyzes the touch pattern CP, but it is an object or a specific effect (function) that is not used in the content game currently provided to the user, the object recognition unit 306d will filter it. can

When the touch pattern CP matches the object (or a specific effect of the game) in the object recognition unit 306d, the determination unit 307 may determine the touch information finally input to the display area DA. . For example, it is possible to match whether the touch information is recognized as the touch pattern CP but is not an object (tool) but touch information by a player's finger.

The determination unit 307 of the multi-table type cognitive training apparatus 100 of the present invention determines when a plurality of players play a content game on the display unit 100a, and the touch information input to the display area DA is determined by which player. It may be determined whether it is input touch information.

The determination unit 307 includes an image separation unit 416 that separates the image captured by the second sensor unit 130b and the fourth sensor unit 130d into an image centered on the player's face and an image centered on the player's motion; , a facial image conversion unit 417 and a motion image conversion unit that converts the facial image and motion image separated by the image separation unit 416 into an image that can be extracted using a learning model 420 by deep learning ( 418), a pattern image conversion unit 419 that converts touch information input to the display unit (a touch contact unit contacted by a touch point of a tool or finger), a face image conversion unit 417, a motion image conversion unit ( 418) and a learning model 420 built by machine learning to extract images for each player based on the transformed image output from the pattern image conversion unit 419, and the player (face), the player's motion (movement), and the player A separation extraction unit 421 for extracting an image of an object (tool: touch information input to the display unit) used by , and a location information providing unit 424 in which the location information of the players provided from the first sensor unit 130a is stored; , an inspection unit 422 that inspects which player the touch information is generated by based on the images extracted from the separation extraction unit 421 and the location information of the players provided from the location information providing unit 424, and the inspection unit ( When the touch information and the players are matched by 422 , a determination unit 423 that determines which player is the touch input for the touch information input for each position of the display area DA is included.

Here, the facial image conversion unit 417 , the motion image conversion unit 418 , and the pattern image conversion unit 419 may perform machine learning on the image captured by the sensor unit and the touch information input to the markup unit. Proceed with conversion.

As shown in FIG. 16 , when four players (player1, player2, player3, player4) play a content game provided by the multi-table type cognitive training apparatus 100, using a hammer each player has The process may be performed by inputting touch information to the touch panel of the display unit 100a. Although the drawing shows the use of a tool (object), this is not fixed. Accordingly, the touch information may be input by a finger touch of each player's finger or the touch information may be input using another tool.

Among the bubbles displayed in the display area, when each player needs to pop their own bubble, the player uses a hammer (tool) to touch the display area where the bubble is displayed. However, if multiple players simultaneously input touch information, or if the bubble to be touched by a player is close to the bubble of an adjacent player, or if a specific area is simultaneously struck with a hammer, which player touches the bubble first? It is necessary to accurately determine whether

To this end, the multi-table type cognitive training apparatus 100 of the present invention may use a learning model 420 built by machine learning. For the machine learning of the learning model 420, not only a neural network, but also a machine learning algorithm such as a decision tree, a Bayesian network, or a support vector machine (SVM) may be used.

For example, it may include neural networks having fully connected layers, such as a Mask Recurrent Convolutional Neural Network (Mask-RCNN). Accordingly, the neural network constituting the learning model 372 may include a feature pyramid network (FPN), a core neural network (Resnet), and a region proposal network (Region Proposal Network).

The learning model 420 includes touch information input on the display area DA by players who play a content game provided by the multi-table type cognitive training apparatus 100 and touch information by a marker M or a finger. It can be built by deep learning learning based on the motion (movement) images for the players and multiple players.

The recognition unit 306 analyzes whether touch information input to the display area DA constitutes the touch pattern CP, and recognizes whether the analyzed touch pattern CP is an object (tool) or a finger of a player. Also, the object recognition unit 306d may recognize whether the touch pattern CP is a function used in a content game. As such, the recognized touch pattern CP is finally determined by the determination unit 307 and controlled through the control unit 305 .

Also, when a plurality of players play a content game, the determination unit 307 may recognize and determine touch information input by each player. Separating, by the determination unit 307, the player image provided from the second sensor unit 130b and the fourth sensor unit 130d into a face image of the player and a motion image of the player; converting the separated face image and motion image of the player into an image capable of machine learning by the learning model 372; converting the touch information input to the display area DA into an image capable of machine learning by the learning model 372; performing machine learning by the learning model 372 on the converted face image, motion image, and touch information image; extracting a related image (a face image, a motion image, and a corresponding touch information image) for each player with respect to the image subjected to machine learning by the learning model 372; examining the respective touch information and corresponding players based on the extracted image and the location information of the players provided by the first sensor unit 130a; and determining a player for each touch information.

As described above, the players corresponding to the plurality of pieces of touch information input to the display area DA determine the positions of the players by the first sensor unit 130a disposed along the periphery of the side surface of the display unit 100a. It can be determined whether it is located on the left or right or up and down (front and rear) of (100a). In addition, when any one of the players located on the left, right or up and down uses a hammer (tool) to pop the bubble in the opposite area, the position of the player who pops the bubble is closer to the display unit 100a than the player located on the opposite side .

If the distance between the pair of players located on the left and right sides or the upper and lower sides and the display unit 100a is similar, the player's face image captured by the second sensor unit 130b and the fourth sensor unit 130d and the player operation The player may be determined by additionally considering the image.

In addition, when at least one or more pieces of touch information are input to a specific area of the display area DA (eg, when a plurality of players touch a specific area with their hammers), a touch input to the display area DA It can be determined in consideration of the touch time of the information, the face of the player, the motion image, and the location information of the player.

As such, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention can provide various cognitive training and cognitive stimulation such as visual, auditory and physical movements to dementia patients or patients with low cognitive ability. can have an effect.

In addition, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention are effective in that a plurality of dementia patients or patients with low cognitive ability can simultaneously receive cognitive training and cognitive stimulation in a cooperative manner. there is

In addition, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention analyze various touch patterns input to the display area by players while playing the content game through a learning model based on deep learning. , there is an effect of accurately recognizing an object or player inputting a touch pattern.

In addition, the multi-table type cognitive training apparatus using artificial intelligence and the player recognition method using the same according to the present invention provide a player providing a touch input in a content game based on the motion of a player who plays the content game and patterns input to the display area have the effect of distinguishing them.

The embodiments according to the present invention described above may be implemented in the form of program instructions that can be executed through various computer components and recorded in a computer-readable recording medium. The computer-readable recording medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the computer-readable recording medium may be specially designed and configured for the present invention, or may be known and used by those skilled in the computer software field. Examples of computer-readable recording media include hard disks, magnetic media such as floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and magneto-optical media such as floppy disks. medium), and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. A hardware device may be converted into one or more software modules to perform processing in accordance with the present invention, and vice versa.

The specific implementations described in the present invention are only examples, and do not limit the scope of the present invention in any way. For brevity of the specification, descriptions of conventional electronic components, control systems, software, and other functional aspects of the systems may be omitted. In addition, the connection or connection members of the lines between the components shown in the drawings exemplarily represent functional connections and/or physical or circuit connections, and in an actual device, various functional connections, physical connections that are replaceable or additional may be referred to as connections, or circuit connections. In addition, unless there is a specific reference such as “essential” or “importantly”, it may not be a necessary component for the application of the present invention.

In addition, although the detailed description of the present invention has been described with reference to a preferred embodiment of the present invention, those skilled in the art or those having ordinary knowledge in the art will appreciate the spirit of the present invention described in the claims to be described later. And it will be understood that various modifications and variations of the present invention can be made without departing from the technical scope. Accordingly, the technical scope of the present invention should not be limited to the contents described in the detailed description of the specification, but should be defined by the claims.

100: multi-table type cognitive training device
100a: display unit
100b: base unit
120: first support
121: second support
123: means of transportation
130a: first sensor unit
130b: second sensor unit
130c: third sensor unit
130d: fourth sensor unit
301: input unit
302: storage
303: driving unit
304: content provider
305: control unit
306: recognition unit
307: judgment unit
308: display unit
TP: touch point
TA: touch area
CP: touch pattern
TD: direction of touchpoint

Claims (6)

a display unit including a touch panel for displaying content games for providing cognitive training or cognitive stimulation to a user and a plate on which the touch panel is seated;
a base unit supporting the display unit; and
A management system for controlling the display unit and the base unit,
The management system is
a display unit for inputting touch information and displaying a content game;
a storage unit for storing information on content games and information on users;
a driving unit that moves the display unit up and down, rotates and tilts;
a content providing unit providing a content game for cognitive training and a content game for cognitive stimulation training to a user;
a recognition unit for recognizing a touch pattern based on the touch information input through the display unit;
a determination unit for determining whether the touch pattern recognized by the recognition unit is touch information input by which user from among a plurality of users;
a first sensor unit disposed along a periphery of a side surface of the display unit to obtain location information of the user;
a second sensor unit disposed on the inside corresponding to the display area of the display unit to obtain face information of the user;
a third sensor unit disposed along a non-display area between the touch panel and the plate of the display unit to obtain a position of a user's marker and information included in the marker; and
The upper space of the display unit includes a fourth sensor unit for acquiring movement information of a plurality of users who play a content game,
The determination unit of the management system uses a learning model built by machine learning based on the touch information input through the display unit and the user's face image and motion image provided by the second sensor unit and the fourth sensor unit. Extracting the user's face and motion image corresponding to the touch information,
Determining a user corresponding to the touch information input to the display unit using the image extracted from the learning model and the user's location information provided from the first sensor unit
Multi-table cognitive training device. According to claim 1,
The judging unit,
an image separation unit separating the images captured by the second sensor unit and the fourth sensor unit into an image centered on the user's face and an image centered on the user's motion;
a facial image conversion unit for converting a face image among the images separated by the image separation unit;
a motion image conversion unit for converting a motion image among the images separated by the image separation unit;
a pattern image conversion unit converting touch information input to the display unit;
a learning model constructed by machine learning to extract a face image and a motion image corresponding to each user's touch information based on the images converted by the face image conversion unit, the motion image conversion unit, and the pattern image conversion unit;
a separation extraction unit for separating and extracting a user, a user's motion, and a tool used by the user from the learning model; and
A determination unit configured to determine a user corresponding to the touch information input to the display unit by using the image extracted from the separation extraction unit and the user's location information by the first sensor unit
Multi-table cognitive training device. 3. The method of claim 2,
The judging unit,
Further comprising an inspection unit that matches the image extracted from the separation extraction unit and the location information provided by the first sensor unit to check which user the touch information is input by
Multi-table cognitive training device. According to claim 1,
The touch information input to the display unit is information input to a touch point of a finger of a user who plays a content game or a marker attached to a tool used to play a content game.
Multi-table cognitive training device. a display unit including a touch panel for displaying content games and a plate on which the touch panel is mounted; A player recognition method using a multi-table type cognitive training apparatus comprising a base unit supporting the display unit and a management system controlling the display unit and the base unit,
The management system includes a display unit for inputting touch information and displaying a content game, a storage unit storing information about content games and information on a user, a driving unit for moving the display unit up and down, rotation and tilt, and to the user. A content providing unit providing a content game for cognitive ability training and a content game for cognitive stimulation training, a recognition unit recognizing a touch pattern based on touch information input through the display unit, and a touch recognized by the recognition unit A determination unit for determining whether the pattern is touch information input by which user from among a plurality of users, a first sensor unit disposed along a periphery of a side surface of the display unit to obtain the user's location information; A second sensor unit disposed inside the display area to obtain face information of the user, and disposed along the non-display area between the touch panel and the plate of the display unit, the location of the user's marker and information contained in the marker A third sensor unit for obtaining a, and a fourth sensor unit for obtaining movement information of a plurality of users who play a content game in the upper space of the display unit,
The judging unit,
separating each user's face image and user's motion image for a plurality of users provided from the second sensor unit and the fourth sensor unit;
converting the separated user's face image and motion images into images that can be learned by a learning model built by machine learning;
converting the touch information input to the display unit into an image that can be learned by a learning model built by machine learning;
After performing machine learning on the converted user's face image, motion image, and touch information images through a learning model, extracting a user face image, a motion image, and a touch information image corresponding thereto for each user; and
After examining each user corresponding to the touch information based on the extracted images and location information on each of the plurality of usage fields provided by the first sensor unit, each user corresponding to the touch information is taking steps to decide
A player recognition method using a multi-table type cognitive training device. 6. The method of claim 5,
The touch information input to the display unit is information input to a touch point of a finger of a user who plays a content game or a marker attached to a tool used to play a content game.
A player recognition method using a multi-table type cognitive training device.

Images