Game table system

Abstract

To provide a novel game table system that uses image recognition to recognize which players have placed what bets at a game table.SOLUTION: A game table system 101 includes: a first association unit 133 that associates chips with players on the basis of the recognized chip positions and the players' bodies; an ID identification unit 134 that identifies the player ID of a player seated at the game table; a combination determination unit 136 that determines a combination of the player ID and a reference image of the player; a second association unit 137 that identifies the player ID of a player depicted in an image on the basis of the determined combinations and associates the player ID with the player depicted in the image; and a chip operation player identification unit 138 that identifies the player ID of a player who has operated a chip on the basis of the association between the player depicted in the image and the player ID, and the association between the player and the chip.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a gaming table system that is applied to a gaming table where games using gaming chips are played. [Background technology]

[0002] Conventionally, systems have been proposed for recognizing chips bet by players at gaming tables where games are played using gaming chips. Systems have also been proposed for recognizing players playing at gaming tables. By linking recognized chips with recognized players, it is possible to recognize the bet amounts of each player, enabling patron ratings and fraud detection. Summary of the Invention [Problem to be solved by the invention]

[0003] However, in conventional systems, when attempting to build a system that recognizes which player bets which gaming chips by photographing the gaming chips betted and the players betting the gaming chips with a camera and performing image recognition, it is necessary to learn images of the players in order to recognize the players.

[0004] It is practically difficult to prepare and train such learning images for all players who may come to a casino and play a game. Even if it were possible to learn images of all players who may come to a casino and play a game, the processing load would be large if player recognition were to be performed on all learned players.

[0005] The present invention has been made in consideration of the above-mentioned problems, and aims to provide a novel gaming table system that uses image recognition to recognize what kind of bets each player has made at a gaming table. [Means for solving the problem]

[0006] One aspect of the gaming table system includes a first camera that photographs chips placed on a gaming table in a casino, chip recognition means that recognizes the positions, types, and number of chips on the gaming table by performing image recognition using AI technology on the image photographed by the first camera, a second camera that photographs a player playing a game at the gaming table to generate an image, body recognition means that recognizes the body of the player playing a game at the gaming table by performing image recognition using AI technology on the image photographed by the second camera, first association means that associates the chips recognized by the chip recognition means with the player whose body is recognized by the body recognition means based on the positions of the chips recognized by the chip recognition means and the body of the player recognized by the body recognition means, ID identification means that identifies a player ID of a player seated at the gaming table, reference image acquisition means that acquires a reference image of the player seated at the gaming table, and a reference image acquisition means that acquires the player ID of the player identified by the ID identification means and the reference image. The system is configured to include a combination determination means for determining a combination with a reference image of the player acquired by the acquisition means, a second association means for specifying the player ID of a player appearing in an image taken by the second camera based on the plurality of combinations determined by the combination determination means for a plurality of players seated at the gaming table, and for associating the specified player ID with the player appearing in the image taken by the second camera, a chip operating player identification means for specifying the player ID of a player who operated the chip based on the association between the player and the player ID by the second association means and the association between the player and the chip by the first association means, and a recording means for storing the player ID specified by the chip operating player identification means in association with the type and number of chips recognized by the chip recognition means, or for recording the player ID specified by the chip operating player identification means in association with a total amount based on the type and number of chips recognized by the chip recognition means.

[0007] In the gaming table system, the ID identifying means may identify the player ID of a player seated at the gaming table by reading the player ID from the member's card of the player.

[0008] In the above gaming table system, the ID identification means may identify the player ID of the player seated at the gaming table by performing image recognition using AI technology on an image of the player seated at the gaming table.

[0009] In the gaming table system described above, the reference image acquisition means may acquire, as the reference image of the player, an image used by the ID identification means to identify the player ID.

[0010] In the above gaming table system, if the player ID of a player seated at the gaming table cannot be identified by the image recognition, the ID identification means may identify the player ID of the player appearing in another image taken by the second camera before or after the image in question.

[0011] The gaming table system may further include a database storing the player IDs of registered players in association with reference images, and the combination determination means may determine the combination of the player ID and the reference image by referring to the database using the player ID identified by the ID identification means.

[0012] In the gaming table system described above, the reference image acquisition means may acquire the reference image of the player by photographing the player whose player ID has been identified by the ID identification means.

[0013] In the above gaming table system, the reference image acquisition means may acquire the reference image of a player by receiving a player's designation from a dealer and cutting out a partial image of the designated player from an image obtained by photographing the player whose player ID has been identified by the ID identification means, and the combination determination means may combine, for the player, the player ID identified by the ID identification means and the reference image of the player acquired by the reference image acquisition means.

[0014] In the above gaming table system, the reference image acquisition means may acquire an image of a player whose player ID has not been associated by the second association means from images obtained by photographing the player whose player ID has been identified by the ID identification means as the reference image of that player, and the combination determination means may combine, for that player, the player ID identified by the ID identification means with the reference image of that player acquired by the reference image acquisition means.

[0015] The gaming table system may further include a database that stores the player ID of a registered player in association with a reference image, and when a player whose player ID and reference image are not stored in the database sits at the gaming table, the ID identification means may identify the player ID of the player by reading the player ID of the player from the player's membership card for new registration, and the reference image acquisition means may acquire the reference image of the player by photographing the player whose player ID is identified by the ID identification means, and the database may store the player ID of the newly registered player identified by the ID identification means and the reference image of the player acquired by the reference image acquisition means in association with each other, and after the player is registered in the database, the combination determination means may refer to the database to determine the combination for a player who sits at any of the gaming tables in the casino.

[0016] In the gaming table system, the reference image acquisition means may acquire a plurality of the reference images for the same player, and the combination determination means may combine a plurality of the reference images for the same player ID.

[0017] In the above gaming table system, the reference image acquisition means may acquire a new reference image of the player after the combination determination means determines a combination of the player ID and the reference image for the player, and the combination determination means may replace the reference image to be combined with the player ID with the new reference image.

[0018] In the gaming table system, the combination determination means may exclude the combination of a player who has left the gaming table from the combinations used for association by the second association means.

[0019] In the above gaming table system, the second matching means may identify players who are likely to be in the casino, and use the combinations of the identified players as candidates to match the identified player IDs with players appearing in images captured by the second camera.

[0020] In the above gaming table system, the second matching means may identify the player ID of the player appearing in the image captured by the second camera based also on the combinations determined by the combination determination means for multiple players seated at other gaming tables within the casino.

[0021] In the above gaming table system, when the second matching means cannot match the identified player ID with the player appearing in the image taken by the second camera, it may use other images taken by the second camera before and after the image in question to match the identified player ID with the player appearing in the other images.

[0022] The above gaming table system may further include a database that stores the player IDs of registered players in association with reference images, and the reference image acquisition means may refer to the database to identify the reference image of the player appearing in the image captured by the second camera.

[0023] The above gaming table system may further include a display means for displaying the reference image identified by the reference image acquisition means, and an input receiving means for receiving input of approval or rejection from the dealer regarding the reference image displayed on the display means, and the reference image acquisition means may acquire the reference image of the player whose rejection input has been received by the input receiving means, and store the reference image in the database in association with the player ID of the player.

[0024] In the above gaming table system, the chip recognition means may include a first chip recognition means that recognizes the position of chips from an image generated by photographing chips on the gaming table, and a second chip recognition means that recognizes the type and number of chips from an image generated by photographing chips on the gaming table.

[0025] In the above gaming table system, the chip recognition means may recognize the position, type, and number of the chip by reading an RFID tag embedded in the chip, the body recognition means may recognize the position of a player playing a game at the gaming table based on the image, and the first association means may associate the player with the chip based on the position of the player recognized by the body recognition means and the position of the chip recognized by the chip recognition means. [Brief explanation of the drawings]

[0026] [Figure 1] FIG. 1 is a diagram showing a gaming table in a casino to which a gaming table system according to an embodiment of the present invention is applied. [Figure 2]FIG. 2 is a block diagram showing the configuration of the gaming table system according to the first embodiment of the present invention. [Figure 3] FIG. 3 is a diagram showing an example of a first image captured by the first camera according to the first embodiment of the present invention. [Figure 4] FIG. 4 is a diagram showing the detection result of the chip stack by the chip recognition unit in the first embodiment of the present invention. [Figure 5] FIG. 5 is a diagram showing an example of a partial image of a chip stack detected by the chip recognition unit according to the first embodiment of the present invention. [Figure 6] FIG. 6 is a diagram showing a result of body recognition by the body recognition unit according to the first embodiment of the present invention. [Figure 7] FIG. 7 is a diagram showing a second image onto which the detection frame of the gaming chip is mapped in the embodiment of the present invention. [Figure 8] FIG. 8 is a diagram showing a result of associating players with player IDs by the second associating unit according to the first embodiment of the present invention. [Figure 9] FIG. 9 is a diagram showing the result of associating stacks of gaming chips with player IDs by the chip-operating player identification unit according to the first embodiment of the present invention. [Figure 10] FIG. 10 is a block diagram showing the configuration of a gaming table system according to the second embodiment of the present invention. [Figure 11] FIG. 11 is a block diagram showing the configuration of a gaming table system according to the third embodiment of the present invention. [Figure 12] FIG. 12 is a block diagram showing the configuration of a gaming table system according to the fourth embodiment of the present invention. [Figure 13] FIG. 13 is a block diagram showing the configuration of a gaming table system according to the fifth embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0027] Hereinafter, embodiments of the present invention will be described with reference to the drawings. Note that the embodiments described below are examples of how the present invention can be implemented, and the present invention is not limited to the specific configurations described below. When implementing the present invention, specific configurations corresponding to the embodiments may be appropriately adopted.

[0028] (First embodiment) Fig. 1 is a diagram showing a gaming table in a casino to which a gaming table system according to a first embodiment of the present invention is applied. As shown in Fig. 1, a dealer 502 is positioned on one side of a gaming table 501, and multiple players 503 are seated at the gaming table 501 facing the dealer 502. At the gaming table 501, a chip tray 504 for storing gaming chips of the dealer 502 is provided in front of the dealer 502, and player positions 505 are provided on the side of each player 503.

[0029] Each player position 505 is provided with a betting area 507 where a player 503 seated at that player position 505 can bet gaming chips (hereinafter simply referred to as "chips") 506. In this embodiment, the gaming table 501 is a gaming table for baccarat, and the betting area 507 of each player position 505 is partitioned into multiple betting targets such as player, banker, tie, etc.

[0030] The gaming table 501 is equipped with a first camera 11 for photographing chips 506 on the gaming table 501, and a second camera 12 for photographing a player 503 and chips 506. The first camera 11 and the second camera 12 are attached to poles provided on the gaming table 501, and photograph the chips 506 on the gaming table 501 from diagonally above.

[0031] The gaming table 501 is also provided with a touch panel 15 for the dealer 502 to perform touch operations.

[0032] 2 is a block diagram showing the configuration of a gaming table system according to a first embodiment of the present invention. As shown in FIG. 2, a gaming table system 101 according to this embodiment includes a first camera 11, a second camera 12, a computing device 13, a card reader 14, a touch panel 15, and a storage device 16. The computing device 13 includes a chip recognition unit 131, a body recognition unit 132, a first association unit 133, an ID identification unit 134, a reference image acquisition unit 135, a combination determination unit 136, a second association unit 137, and a chip-operating player identification unit 138. Note that each function of the computing device 13 may be realized by a computer including a processor, a memory, and a storage device executing a computer program according to this embodiment.

[0033] The first camera 11 photographs the chips 506 placed on the gaming table 501 to generate a first image. The first camera 11 photographs the table surface of the gaming table 501, which has a plurality of betting areas 507, from diagonally above. As a result, the first image shows the plurality of betting areas 507 in the depth direction when viewed from the first camera 11, and also shows a stack of chips 506 that is relatively close and has a large downward angle, and a stack of chips 506 that is relatively far away and has a small downward angle. The first camera 11 may continuously photograph to generate successive images, or may photograph and generate images in response to a predetermined trigger.

[0034] The second camera 12 captures an image of a player 503 playing a game on the gaming table 501 and chips 506 placed on the gaming table 501 to generate a second image. The imaging direction and angle of view of the second camera 12 are set so that the face of the player 503 playing a game on the gaming table 501 is captured in the image.

[0035] Although first camera 11 and second camera 12 are installed on different poles, separated into left and right, first camera 11 and second camera 12 may be installed one above the other on the same pole.

[0036] The card reader 14 is operated by a dealer 502 to read a player ID from a member's card received from a player 503 seated at a gaming table 501. A member's card is issued to a player 503 who has registered as a member, and the player ID is stored in the member's card.

[0037] The touch panel 15 is installed in front of the dealer 502 so that the dealer 502 can check various pieces of information. In the example of FIG. 1, the touch panel 15 is embedded in the gaming table 501, but instead, it may be installed upright on the gaming table 501. The touch panel 15 has a function of accepting touch input, and in particular, as will be described later, accepts touch input from the dealer 502 when the dealer 502 designates a player 503 whose membership card has been read. In other words, the touch panel 15 functions as a display means as well as an input accepting means.

[0038] 3 is a diagram showing an example of a first image captured by the first camera in the first embodiment of the present invention. As shown in Fig. 3, the first image 110 shows multiple players 503 seated at a gaming table 501, as well as stacks of chips 506 bet by the players 503 in a betting area 507.

[0039] As described above, first camera 11 photographs multiple betting areas 507 from diagonally above, so that the top surface of the top chip 506 in the chip stack and the sides of all chips 506 in the chip stack are visible in first image 110. Also visible in first image 110 are relatively close betting areas 507, relatively far betting areas 507, and chips 506 bet on those betting areas 507. Chips 506 in relatively far chip stacks appear relatively small, and chips 506 in relatively close chip stacks appear relatively large.

[0040] The chip recognition unit 131 recognizes the position, type, and number of chips 506 on the gaming table 501 by performing image recognition using AI technology on the first image 110 obtained by capturing an image with the first camera 11. To this end, the chip recognition unit 131 includes a first chip recognition unit that recognizes the position of the chips 506 from the first image 110, and a second chip recognition unit that recognizes the type and number of chips 506 from the first image 110. The first chip recognition unit detects the chip stack from the first image 110, thereby identifying the position of the chip stack on the gaming table 501.

[0041] 4 is a diagram showing the detection result of a chip stack by the chip recognition unit in the first embodiment of the present invention. As shown in FIG. 4, the first chip recognition unit of the chip recognition unit 131 cuts out a partial image of the chip stack by surrounding the area of ​​the chip stack in the first image 110 with a rectangular detection frame. If the chip stack is in the bet area 507, the first chip recognition unit determines which specific bet object in the bet area 507 the chip stack is in. Specifically, the first chip recognition unit estimates the center position of the bottom chip 506 of the detected chip stack, for example, and determines which bet object area on the gaming table 501 the center position corresponds to.

[0042] FIG. 5 is a diagram showing an example of a partial image of a chip stack detected by the chip recognition unit of the first embodiment of the present invention. The chip 506 of this embodiment has a center line on its side. The color of the center line represents the type (price) of the chip 506. The second chip recognition unit of the chip recognition unit 131 then detects the center lines from this partial image of the chip stack and determines their color. As a result, the second chip recognition unit determines the number of chips 506 based on the number of detected center lines, and determines the type of each chip 506 based on the color of the detected center line.

[0043] In this embodiment, as described above, the chip recognition unit 131 identifies the position of the stack of chips 506 using the first chip recognition unit, and identifies the type and number of chips 506 in each stack using the second chip recognition unit. However, instead, the chip recognition unit 131 may use one recognition engine to identify the position of the stack of gaming chips 506 and also identify the type and number of chips 506 in that stack.

[0044] Body recognition unit 132 recognizes the body of player 503 playing a game on gaming table 501 by performing image recognition using AI technology on second images obtained by second camera 12. Second camera 12 continuously takes images at a predetermined frame rate to generate successive second images arranged in chronological order, and body recognition unit 132 performs image recognition on each of these successive second images to recognize the body of player 503.

[0045] Fig. 6 is a diagram showing the results of body recognition by the body recognition unit of the first embodiment of the present invention. For convenience, Fig. 6 illustrates second image 120 having the same content as first image 110. However, as described above, first camera 11 and second camera 12 are installed in different positions and directions, and therefore first image 110 and second image 120 do not actually have the same content.

[0046] The body recognition unit 132 performs pose tracking on the second image 120 to detect predetermined characteristic body parts (face (center of both eyes), both shoulders, both elbows, both wrists, and fingers of both hands) of each player 503, and generates a skeletal model of the player 503. In FIG. 6, the black dots are the detected characteristic body parts of the player 503, and a skeletal model is generated by connecting the black dots for each player 503. Any existing technology can be used for such pose tracking. The body recognition unit 132 further detects the face of the player 503 as a body part of the player 503, surrounds it with a rectangular detection frame, and combines the detection frame and the skeletal model for each player 503 to recognize it as the body of the player 503.

[0047] The first association unit 133 associates the chip 506 recognized by the chip recognition unit 131 with the player 503 whose body is recognized by the body recognition unit 132, based on the position of the chip 506 recognized by the chip recognition unit 131 and the body of the player 503 recognized by the body recognition unit 132. To this end, the first association unit 133 maps a detection frame of the chip 506 detected from the first image 110 onto the second image 120 in which the body is recognized. This mapping may be performed based on the alignment of the first camera 11 and the second camera 12, or, for example, mapping may be performed based on a bet area 507 in the image. Conversely, the first association unit 133 may map a detection frame in which a skeletal model and a face recognized in the second image 120 are detected onto the first image 110.

[0048] 7 is a diagram showing a second image onto which the detection frame of a gaming chip is mapped in an embodiment of the present invention. The first association unit 133 maps the detection frame of the chip 506 at the timing when the chip 506 appears in the first image 110 from the first images 110 captured continuously in chronological order onto the second image 120 at that timing. Alternatively, the first association unit 133 may map the detection frame of the chip 506 detected in the first image 110 onto the second image 120 at that timing when the position of the detection frame of the chip 506 no longer changes.

[0049] The first association unit 133 associates the player 503 having the finger closest to the detection frame of the chip 506 with the detection frame of the chip 506 in the second image 120 onto which the detection frame of the chip 506 is mapped. In this way, the first association unit 133 associates the chip 506 recognized by the chip recognition unit 131 with the player 503 whose body is recognized by the body recognition unit 132. At this point, the contents of the chip 506 (position, type, and number) and the player 503 are associated, but this is merely an association between the two in the image, and it has not yet been determined who the player 503 shown in the image is, i.e., the player ID of the player 503.

[0050] It should be noted that the stacks of the other chips 506 shown in FIG. 7 are associated in frames with different timings.

[0051] The ID identification unit 134 identifies the player ID of the player 503 seated at the gaming table 501. In this embodiment, the ID identification unit 134 reads the player ID of the player 503 from the member's card of the player 503 seated at the gaming table 501 using the card reader 14, and acquires the read result from the card reader 14, thereby identifying the player ID.

[0052] The reference image acquisition unit 135 acquires a reference image of the player 503 seated at the gaming table 501. In this embodiment, the reference image acquisition unit 135 acquires a partial image of the face detected by the body recognition unit 132 and sets this as the reference image of the player 503.

[0053] The combination determination unit 136 determines a combination of the player ID of the player 503 identified by the ID identification unit 134 and the reference image of the player 503 acquired by the reference image acquisition unit 135. As shown in FIG. 6, if there are multiple faces in the second image 120, the body recognition unit 132 detects multiple faces. Therefore, in this state, it is not possible to know which player 503's face is detected as the player ID read by the card reader 14.

[0054] Therefore, when the dealer 502 reads the player ID from the member's card using the card reader 14, the touch panel 15 synchronizes with this and displays the screen shown in Fig. 6, which shows the face detection results from the second image 120. In response to the face detection results displayed on the touch panel 15, the dealer 502 taps the face of the player 503 whose member's card has been read, thereby specifying the player 503 corresponding to the read player ID (i.e., the player 503 who has newly arrived at the gaming table 501).

[0055] The combination determination unit 136 receives this designation from the touch panel 15 and determines a combination of the player ID of the player 503 read from the member's card and the face image (reference image) of the player 503.

[0056] The second association unit 137 identifies the player ID of the player 503 appearing in the second image 120 captured by the second camera 12 based on multiple combinations of player IDs and reference images determined by the combination determination unit 136 for multiple players 503 seated at the gaming table 501, and associates the identified player ID with the player 503 appearing in the second image 120 captured by the second camera 12.

[0057] 8 is a diagram showing the result of associating a player with a player ID by the second associating unit in the first embodiment of the present invention. As shown in FIG. 8, the second associating unit 137 associates the player 503 recognized in the second image 120 with the player ID. To this end, when the body recognition unit 132 detects the face from the second image 120 to obtain a partial image of the face (face image), the second associating unit 137 selects a reference image that is closest to this face image from among the multiple reference images acquired by the reference image acquisition unit 135 for the gaming table 501. Then, the second associating unit 137 associates the player ID combined with the selected reference image with the face in the partial image.

[0058] The chip operating player identification unit 138 identifies the player ID of the player 503 who operated the gaming chip 506, based on the association between the player 503 and the player ID made by the second association unit 137 and the association between the player 503 and the chip 506 made by the first association unit 133. As described above, the first association unit 133 associates the chip 506 with the player 503 whose finger is closest to the chip 506 at the time the chip 506 is detected (see FIG. 7), and the second association unit 137 associates the player ID with the player 503 in the second image 120 (see FIG. 8).

[0059] 9 is a diagram showing the result of associating a stack of gaming chips with a player ID by the chip operating player identification unit in the first embodiment of the present invention. The chip operating player identification unit 138 associates the stack of chips 506 with the player ID based on the associations made by the first association unit 133 and the second association unit 137, and identifies the player ID of the player 503 who operated (placed) the stack of chips 506. This makes it possible to identify which player 503 bet the stack of chips 506 recognized on the gaming table 501.

[0060] The storage device 16 stores the player ID identified by the chip operating player identification unit 138 and the type and number of chips 506 recognized by the chip recognition unit 131 in association with each other. Note that, for the chips 506, the storage device 16 may record the total amount based on the type and number of chips in association with the player ID instead of or in addition to the type and number of chips. This makes it possible to grasp the amount of bet by each player 503 on the gaming table 501, which can be useful for patron rating and fraud detection.

[0061] (Second embodiment) FIG. 10 is a block diagram showing the configuration of a gaming table system according to a second embodiment of the present invention. In this embodiment, the gaming table system 102 identifies the player ID of a player 503 seated at a gaming table 501 through facial recognition. That is, in this embodiment, the ID identification unit 134 identifies the player ID of the player 503 seated at the gaming table 501 by performing image recognition using AI technology on an image of the player 503 seated at the gaming table 501. For this purpose, the ID identification unit 134 has a facial recognition engine that has learned pre-registered facial images of the player 503. For example, deep learning technology or SVM (support vector machine) is used as the AI ​​technology for this facial recognition engine. The image used for facial recognition by the ID identification unit 134 may be captured by the second camera 12.

[0062] The reference image acquisition unit 135 acquires the face image captured by the second camera 12 and used for face authentication as a reference image. As in the first embodiment, the second association unit 137 selects the closest reference image for the face image extracted from the second image 120 captured by the second camera 12, thereby matching the reference image with the face image. However, a face image that is not matched with any reference image is one for which a reference image has not yet been prepared. Therefore, the ID identification unit 134 performs processing to identify a player ID by face authentication for face images that have not been associated by the second association unit 137.

[0063] According to this embodiment, it is necessary to learn the face image of the player 503 in advance, but in actual operation, the player 503 does not need to present the member's card to the dealer 502 when he arrives at the gaming table 501, and the dealer 502 does not need to scan the member's card with a card reader, which makes the operation more efficient. In other words, in this embodiment, instead of or in addition to issuing a member's card, a photograph of the player 503 for face authentication may be taken.

[0064] (Third embodiment) 11 is a block diagram showing the configuration of a gaming table system according to a third embodiment of the present invention. In this embodiment, the gaming table system 103 is provided with a reference image database 17. The reference image database 17 stores the player ID of a registered player 503 and a reference image associated with the player ID. In the first and second embodiments, the reference image acquisition unit 135 acquired a reference image of the player 503 seated at the gaming table 501. In this embodiment, however, the player 503 is registered, and the reference image acquisition unit 135 acquires a reference image of a newly registered player 503.

[0065] Furthermore, multiple reference images may be acquired for each player 503 to be registered. That is, in the reference image database 17, multiple reference images taken at different angles may be associated with one player 503.

[0066] Then, the combination determination unit 136 determines a combination of the player ID and the reference image by referring to the reference image database 17 using the player ID identified by the ID identification unit 134 and identifying a reference image associated with the player ID. Note that in this embodiment, the ID identification unit 134 may identify the player ID by reading a member's card with the card reader 14 as in the first embodiment, or may identify the player ID by face authentication as in the second embodiment.

[0067] (Fourth embodiment) Fig. 12 is a block diagram showing the configuration of a gaming table system according to a fourth embodiment of the present invention. In the above-described embodiment, a camera for capturing a first image 110 for recognizing the type and number of chips 506 and a camera for capturing a second image 120 for pose tracking of a player 503 are installed as separate cameras, but in the gaming table system 104 of this embodiment, these may be combined into a single camera (second camera 12 in the example of Fig. 12).

[0068] In this case, the chip recognition unit 131 recognizes the chip 506 by performing image recognition on the second image 120 captured by the second camera 12, and the body recognition unit 132 also performs image recognition on the same second image 120 to generate a skeletal model and detect the facial area.

[0069] In the above embodiment, the first camera 11 is set to a relatively narrow angle of view so as to photograph the chips 506 placed in the betting area 507 of the gaming table 501, and the second camera 12 is set to a relatively wide angle of view so as to adequately capture the face of the player 503 and his hand stretched out over the betting area 507. In this embodiment, a relatively wide angle of view is also set so as to adequately capture the face of the player 503, his hand stretched out over the betting area 507, and the chips 506 placed in the betting area 507. In this embodiment, it is desirable that the number of pixels of the second camera 12 is sufficiently large so that the chip recognition unit 131 can adequately recognize the chips 506 even with such a wide angle of view.

[0070] (Fifth embodiment) 13 is a block diagram showing the configuration of a gaming table system according to a fifth embodiment of the present invention. In the above embodiment, a camera photographs chips 506 placed in a betting area 507 to generate an image, and a chip recognition unit 131 performs image recognition on the photographed image to recognize the position, type, and number of chips 506. In this embodiment, an RFID tag capable of identifying information about the type of chip 506 is embedded in the chip 506, and the gaming table system 105 is provided with a betting area reader 18 in the betting area 507 for reading the RFID tag of the chip 506 placed in the betting area 507.

[0071] The betting area reader 18 comprises an antenna installed at each betting object and a reader that controls the antenna. This makes it possible to read the RFID tag of the chip 506 at each player position 505. The chip recognition unit 131 recognizes the position of the chip 506 for each antenna that reads the RFID tag of the chip 506, i.e., which player position 505 and which betting object the chip 506 is located at. In addition, the chip recognition unit 131 recognizes the type and number of chips 506 placed at each betting object based on the type information of the chip 506 stored in the RFID tag read from each betting object by the betting area reader 18 and the number of chips 506.

[0072] In this embodiment, an antenna is provided for each bet object to read the chips 506 placed on each bet object, but an antenna may be provided for each bet area 507. Even with this configuration, it is possible to recognize the type and number of chips 506 bet on each bet area 507, and the bet amount of each player 503.

[0073] In this embodiment, the body recognition unit 132 recognizes the body of the player 503 as in the above-described embodiment, and also recognizes the player position 505 of the player 503 whose body has been recognized. Then, the first association unit 133 associates the player 503 recognized by the body recognition unit 132 with the gaming chip 506 recognized by the chip recognition unit 131, based on the player position 505 recognized by the body recognition unit 132 and the position of the gaming chip 506 recognized by the chip recognition unit 131.

[0074] (Sixth embodiment) Similar to the gaming table system 103 (see FIG. 11), the gaming table system of the sixth embodiment includes a reference image database 17 that stores the player ID of a registered player 503 in association with a reference image. In this embodiment, when a player 503 whose player ID and reference image are not stored in the reference image database 17 arrives at the gaming table 501, the ID identification unit 134 identifies the player ID of the player 503 by reading the player ID of the player 503 from the member's card of the player 503 for new registration.

[0075] When ID identification unit 134 identifies the player ID of newly registered player 503, reference image acquisition unit 135 acquires a reference image of player 503 by photographing player 503 with second camera 12. Reference image database 17 then updates reference image database 17 by storing the player ID of newly registered player 503 identified by ID identification unit 134 and the reference image of player 503 acquired by reference image acquisition unit 135 in association with each other.

[0076] In addition, in this embodiment, after the player 503 is registered in the reference image database 17, the combination determination unit 136 refers to the reference image database 17 and determines a combination of the player ID and the reference image for the player 503 who is seated at any gaming table 501 in the casino.

[0077] According to this embodiment, there is no need to take a photograph of the face of a player 503 playing at a casino in order to register the face in the reference image database. The player 503 can be registered in the reference image database when the player 503 sits at the gaming table 501 for the first time after the casino has installed a gaming table system.

[0078] (Seventh embodiment) The gaming table system of the seventh embodiment, like the gaming table system 103 (see FIG. 11), includes a reference image database 17 that stores the player ID of the registered player 503 in association with a reference image. In this embodiment, the reference image acquisition unit 135 refers to the reference image database 17 to identify the reference image of the player 503 that appears in the image captured by the second camera 12.

[0079] The touch panel 15 displays the reference image specified by the reference image acquisition unit 135, and also receives input of approval or rejection of the displayed reference image from the dealer 502. The dealer 502 compares the reference image displayed on the touch panel 15 with the face of the player 503 actually at the gaming table 501, and inputs approval if he / she recognizes that they are the same person, or inputs rejection if he / she recognizes that they are different people.

[0080] If the reference image acquisition unit 135 receives an approval input on the touch panel 15, the combination determination unit 136 determines a combination of this reference image and a player ID. If the reference image acquisition unit 135 receives a rejection input on the touch panel 15, the reference image acquisition unit 135 acquires a reference image of the player 503 by photographing the player 503 with the second camera 12, and stores the reference image in the reference image database 17 in association with the player ID of the player 503. Then, the combination determination unit 136 determines a combination of this reference image and a player ID.

[0081] The dealer 502 may input a refusal if he / she determines that the reference image displayed on the touch panel 15 is not of sufficient quality to be used as a reference image. Furthermore, even if the dealer 502 recognizes that the reference image and the actual player 503 are the same person, he / she may input a refusal if the appearance of the reference image and the actual player 503 have changed significantly. In short, the dealer 502 can input a refusal if he / she determines that the reference image in the reference image database 17 should be updated.

[0082] According to this embodiment, when a reference image of the player 503 is not registered in the reference image database 17, or when a reference image of the player 503 is registered in the reference image database 17 but it is desired to update the reference image, an image of the player 503 taken on the spot can be registered (updated) in the reference image database 17 as a new reference image of the player 503.

[0083] (Eighth embodiment) In the first embodiment, the first associating unit 133 associates the player 503 whose body is recognized by the body recognition unit 132 with the chip 506 by mapping the detection frame of the chip 506 at the timing when the chip 506 appears in the first image 110, which are taken continuously in chronological order, onto the second image 120 at that timing. Then, the second associating unit 137 associates the player 503 appearing in the second image 120 with the player ID by determining whether the facial image of the player 503 appearing in the second image 120 corresponds to any of the reference images determined by the combination determination unit 136.

[0084] However, the second image 120 is not necessarily an image suitable for matching with the reference image. For example, if the face of the player 503 who operated the chip 506 is not facing forward or is not clearly visible in the second image 120 at the time when the chip 506 appears in the first image 110, the second association unit 137 may not be able to match the second image 120 with the reference image.

[0085] In this case, the second association section 137 may perform matching with the reference image using the second image 120 of the frame before or after the second image 120 at the timing when the chip 506 appears in the first image 110, rather than the second image 120 at the timing when the chip 506 appears in the first image 110. For this reason, the second association section 137 can maintain the association between the chip 506 and the body associated by the first association section 133 by tracking the body recognized by the body recognition section 132 from the second image 120 at the timing when the chip 506 appears in the first image 110 to the second image 120 of that frame.

[0086] According to this embodiment, even if the face of the player 503 who operated the chip 506 is not shown in the second image 120 at the time when the chip 506 appears in the first image 110 in a manner that allows the second association unit 137 to make an association, the second association unit 137 can identify the player ID of the player 503 who operated the chip 506.

[0087] (About the reference image) In the gaming table system 101 of the first embodiment, a reference image of a player 503 seated at a gaming table 501 is prepared by the reference image acquisition unit 135, and thereafter, when the player 503 plays at the gaming table 501, the player ID of the player 503 is identified by the second association unit 137 using the reference image that was initially prepared. As a result, it is not necessary to identify the player 503 from all registered players 503 as candidates every time the player 503 plays a game, and the player 503 can be identified by selecting the relevant player 503 from the players 503 prepared as reference images at the gaming table 501.

[0088] In this way, the reference image of the player 503 is prepared when the player 503 sits at the gaming table 501, but the most appropriate reference image is not necessarily obtained at this timing. Therefore, the reference image acquisition unit 135 may update the reference image by acquiring a face image of the player 503 obtained by photographing the player 503 at any timing after the player 503 starts playing the game at the gaming table 501 as a new reference image instead of or in addition to the original reference image. In other words, the reference image acquisition unit 135 may acquire a new reference image of the player 503 after the combination determination unit 136 has once determined a combination of a player ID and a reference image for the player 503.

[0089] When the reference image acquisition unit 135 updates the reference image, the combination determination unit 136 replaces the reference image to be combined with the player ID with the updated reference image, and determines the combination.

[0090] Furthermore, the combination determination unit 136 excludes combinations of players 503 who have left the gaming table 501 from the combinations used for association by the second association unit 137. Specifically, the combination determination unit 136 excludes from the combinations reference images that have not been associated for a predetermined number of consecutive games by the second association unit 137. This prevents reference images that have become unnecessary because the player 503 has left the gaming table 501 from remaining as candidates for association, thereby reducing the processing load of association by the second association unit 137 and improving the processing speed.

[0091] Conversely, the second associating unit 137 may also identify the player ID of the player 503 depicted in the second image 120 based on the combinations determined by the combination determination unit 136 for multiple players 503 seated at other gaming tables 501 in the casino. This configuration reduces the processing load because it is not necessary to prepare a new reference image every time the player 503 seats at a new gaming table 501. Furthermore, by updating the reference image as described above, an appropriate reference image obtained when the player 503 was at another gaming table 501 can be used.

[0092] When identifying the player ID of the player 503 of the second image 120, the second association unit 137 may perform association by using all players 503 whose reference images have been obtained at any gaming table 501 in the casino as candidates. Alternatively, at each gaming table 501, when the player 503 leaves the gaming table 501, the combination determination unit 136 may exclude the combination of the player 503 from the candidates to be used for association by the second association unit 137 and store the combination for a certain period of time, so that when the player 503 arrives at another gaming table 501, the other gaming table 501 may search for a corresponding combination from the combinations stored for the other gaming table 501 and use the combination at the other gaming table 501.

[0093] (Other variations) The functions of the arithmetic unit 13 may be implemented in multiple distributed computers, or some or all of the functions may be located on a cloud. Furthermore, a computer program for configuring the functions of the arithmetic unit 13 may be installed on a computer via a network or a non-transitory storage medium. In the above embodiment, the gaming table system of the present invention has been described using a baccarat table as an example, but the gaming table system of the present invention can also be applied to gaming tables other than baccarat tables, such as blackjack gaming tables and roulette tables.

[0094] In the above embodiment, the chip 506 has a center line on its side that indicates its type by color, and the chip recognition unit 131 identifies the type of the chip 506 based on the color of the center line. However, instead of or in addition to this, code information representing identification information that uniquely identifies the gaming chip 506 may be written on the side of the chip 506. In this case, the chip recognition unit 131 may identify the identification information of the chip 506 instead of or in addition to the type. Furthermore, the gaming table system may be provided with a chip database, and the type of the chip 506 may be identified based on the identification information of the chip 506.

[0095] Furthermore, the gaming table system of the above embodiment identifies the player 503 who placed (bet) the chips 506 in the betting area 507. However, instead of or in addition to this, the player 503 who removed the chips 506 from the betting area 507 may be identified. In this case, the first association unit 133 maps the detection frame of the chips 506 detected in the first image 110 obtained successively at a timing immediately before the gaming chips 506 disappear from the first images 110 onto the second image 120. Alternatively, when the position of the stack of chips 506 recognized in the betting area 507 moves, the detection frame of the chips 506 detected in the first image 110 at a timing immediately before the movement may be mapped onto the second image 120.

[0096] With this configuration, when the chips 506 paid out to the player 503 who won the game and the chips 506 that the player 503 bet on are collected by the player 503, the player ID of the player 503 who performed such an operation on such chips 506 can be identified.

[0097] In the first embodiment described above, a screen showing the result of face detection from the second image 120 is displayed on the touch panel 15, and the combination determination unit 136 receives from the dealer 502 the designation of a player 503 who has newly arrived at the gaming table 501, and determines a combination of the player ID and the face image of the player 503, which serves as a reference image. In addition to this function, the combination determination unit 136 may have a function of issuing an alert when the dealer 502 makes a mistake in designating the face of the player 503.

[0098] In this case, the gaming table system 101 includes a reference image database 17 that stores correspondence between the faces of registered players 503 and their player IDs. The second association unit 137 executes a warning measure when there is a low degree of match between the face of the player 503 designated by the dealer 502 on the touch panel 15 and the face stored in the reference image database 17 as the face of the player 503 corresponding to the player ID read by the card reader 14. The warning measure may, for example, be to display a message on the touch panel 15 indicating that there is a low degree of match.

[0099] Furthermore, the gaming table system 103 of the third embodiment includes a reference image database 17, and the combination determination unit 136 determines a combination of a player ID and a reference image by referring to the reference image database 17 using the player ID identified by the ID identification unit 134 and identifying a reference image associated with the player ID. Alternatively, the combination determination unit 136 may use the reference image associated with the identified player ID and select a facial image similar to the reference image from among the plurality of facial images extracted from the second image 120. In this selection, the combination determination unit 136 may use a learning-based facial recognition model using AI technology such as deep learning or SVM.

[0100] At this time, the touch panel 15 displays the selected facial image and receives approval for the selected facial image from the dealer 502. The combination determination unit 136 may combine the facial image approved by the dealer 502 with the specified player ID as a reference image. If the facial image selected by the combination determination unit 136 is that of the player 503 who has newly arrived at the gaming table 501, the dealer 502 inputs approval on the touch panel 15, and otherwise inputs a denial of approval.

[0101] If a refusal is input from the dealer 502, the combination determination unit 136 displays the face image with the next highest score in the face recognition as a candidate on the touch panel 15. If a refusal is input, the combination determination unit 136 may perform reinforcement learning on the face recognition model using the result of this face recognition as a negative example.

[0102] Furthermore, the gaming table system 103 of the third embodiment includes the reference image database 17, and the combination determination unit 136 determines a combination of a player ID and a reference image by referring to the reference image database 17 using the player ID identified by the ID identification unit 134 and identifying a reference image associated with the player ID. In this case, the combination determination unit 136 may not directly adopt the reference image identified by referring to the reference image database 17, but may display the reference image on the touch panel 15 as a reference image candidate and combine the reference image with the player ID with the approval of the dealer 502.

[0103] Furthermore, in the gaming table system 102 of the second embodiment, the ID identification unit 134 identifies the player ID of the player 503 seated at the gaming table 501 by performing image recognition using AI technology on an image of the player 503 seated at the gaming table 501. A reference image database 17 may be added to this gaming table system 102. The ID identification unit 134 may extract multiple player ID candidates by image recognition, rather than identifying one player ID by image recognition.

[0104] The combination determination unit 136 refers to the reference image database 17 using the multiple player ID candidates identified by the ID identification unit 134, and extracts multiple reference image candidates from the reference image database 17. The touch panel 15 displays the multiple extracted reference images and accepts a selection from the dealer 502. The dealer 502 compares the multiple reference image candidates displayed on the touch panel 15 with the face of the player 503 who has actually just sat at the gaming table 501, and selects a reference image on the touch panel 15.

[0105] According to this configuration, even if the accuracy of face recognition for identifying the player ID of player 503 is relatively low, it is possible to reduce the possibility of an error in identifying the player ID due to an error in face recognition.

[0106] Furthermore, the second associating unit 137 may identify a player 503 who is likely to be in the casino, and may use combinations of the identified player 503 as candidates to associate the identified player ID with the player 503 who appears in the image captured by the second camera 12. The player 503 who is likely to be in the casino may be, for example, a player 503 who is recognized at any gaming table 501 in the casino within a predetermined time, or a player 503 whose entry at the casino gate is confirmed within a predetermined time.

[0107] Furthermore, in the second embodiment, the ID identification unit 134 identifies the player ID of the player 503 seated at the gaming table 501 by performing image recognition on the image captured by the second camera 12. However, if the quality of the image used for this image recognition is poor and the player ID cannot be identified by image recognition on this image, the ID identification unit 134 may use other images captured by the second camera 12 before and after the image in question to perform image recognition on the other images, and identify the player ID of the player 503 appearing therein.

[0108] Possible cases where the image quality is poor include when an image of the face of the player 503 from the front cannot be captured, when the face of the player 503 is not in focus, or when an obstacle is between the second camera 12 and the face of the player 503 causing occlusion. In these cases, it is desirable to perform image recognition using another image captured at a slightly different time, in the hope that the cause of the quality degradation has disappeared, and to identify the player ID. [Explanation of symbols]

[0109] 11 Camera 1 12 Second Camera 13 Arithmetic unit 14 Card Reader 15 Touch Panel 16 Storage device 17 Reference Image Database 18 Betting Area Leader 101~105 Gaming Table System 110 First Image 120 2nd image 131 Chip recognition unit 132 Body Recognition Department 133 First Correspondence Section 134 ID identification part 135 Reference image acquisition unit 136 Combination determination unit 137 Second Correspondence Section 138 Chip operation player identification unit 501 Gaming Table 502 Dealer 503 players 504 Chip Tray 505 Player Position 506 Gaming Chip 507 Betting Area

Claims

1. a first camera for photographing chips placed on a gaming table in a casino; chip recognition means for recognizing the position, type, and number of chips on the gaming table by performing image recognition using AI technology on the image captured by the first camera; a second camera that captures an image of a player playing a game at the gaming table and generates an image; a body recognition means for recognizing the body of the player playing the game at the gaming table by performing image recognition using AI technology on the image captured by the second camera; a first association means for associating the chip recognized by the chip recognition means with the player whose body is recognized by the body recognition means, based on the position of the chip recognized by the chip recognition means and the body of the player recognized by the body recognition means; ID identification means for identifying a player ID of a player seated at the gaming table; a reference image acquisition means for acquiring a reference image of a player seated at the gaming table; a combination determination means for determining a combination of the player ID of the player identified by the ID identification means and the reference image of the player acquired by the reference image acquisition means; a second association means for identifying the player IDs of the players appearing in the images captured by the second camera based on the plurality of combinations determined by the combination determination means for the plurality of players seated at the gaming table, and for associating the identified player IDs with the players appearing in the images captured by the second camera; a chip operating player specifying means for specifying a player ID of a player who operates a chip based on the association between a player and a player ID by the second association means and the association between the player and the chip by the first association means; a recording means for storing the player ID specified by the chip operating player specifying means and the type and number of chips recognized by the chip recognition means in association with each other, or for recording the player ID specified by the chip operating player specifying means and a total amount based on the type and number of chips recognized by the chip recognition means in association with each other; A gaming table system equipped with:

2. 2. The gaming table system according to claim 1, wherein the ID specifying means specifies the player ID of the player seated at the gaming table by reading the player ID from a member's card of the player.

3. 2. The gaming table system according to claim 1, wherein the ID identification means identifies the player ID of the player seated at the gaming table by performing image recognition using AI technology on an image of the player seated at the gaming table.

4. 4. The gaming table system according to claim 3, wherein the reference image acquisition means acquires, as the reference image of the player, an image used by the ID identification means to identify the player ID.

5. 4. The gaming table system of claim 3, wherein when the player ID of a player seated at the gaming table cannot be identified by the image recognition, the ID identification means uses other images taken by the second camera before and after the image to identify the player ID of the player appearing in the other images.

6. The game device further includes a database storing the player IDs of registered players and reference images in association with each other; 2. The gaming table system according to claim 1, wherein the combination determination means determines a combination of the player ID and the reference image by referring to the database using the player ID identified by the ID identification means.

7. 2. The gaming table system according to claim 1, wherein the reference image acquisition means acquires the reference image of the player by photographing the player whose player ID has been identified by the ID identification means.

8. the reference image acquisition means, upon receiving a designation of a player from a dealer, extracts a partial image of the designated player from an image obtained by photographing the player whose player ID has been identified by the ID identification means, thereby acquiring the reference image of the designated player; 8. The gaming table system according to claim 7, wherein the combination determination means combines, for the player, the player ID identified by the ID identification means with the reference image of the player acquired by the reference image acquisition means.

9. the reference image acquisition means acquires, as the reference image of the player, an image of a player for which association by the second association means has not been performed, from among images obtained by photographing the player whose player ID has been identified by the ID identification means; 8. The gaming table system according to claim 7, wherein the combination determination means combines, for the player, the player ID identified by the ID identification means with the reference image of the player acquired by the reference image acquisition means.

10. The game device further includes a database storing the player IDs of registered players and reference images in association with each other; the ID specifying means, when a player whose player ID and reference image are not stored in the database sits at the gaming table, specifies the player ID of the player by reading the player ID of the player from the member's card of the player for new registration; the reference image acquisition means acquires the reference image of the player by photographing the player whose player ID has been identified by the ID identification means; the database stores the player ID of the newly registered player identified by the ID identification means and the reference image of the player acquired by the reference image acquisition means in association with each other; 2. The gaming table system according to claim 1, wherein the combination determination means, after the player is registered in the database, determines the combination for a player seated at any of the gaming tables in the casino by referring to the database.

11. the reference image acquisition means acquires a plurality of the reference images for the same player; 2. The gaming table system according to claim 1, wherein the combination determination means combines a plurality of the reference images for the same player ID.

12. the reference image acquisition means acquires a new reference image of the player after the combination determination means determines a combination of the player ID and the reference image for the player; 2. The gaming table system according to claim 1, wherein the combination determination means replaces the reference image to be combined with the player ID with the new reference image.

13. 2. The gaming table system according to claim 1, wherein the combination determination means excludes the combination of a player who has left the gaming table from the combinations used for association by the second association means.

14. 2. The gaming table system according to claim 1, wherein the second association means identifies players who are likely to be in the casino, and associates the identified player ID with the player appearing in the image captured by the second camera, using the combination of the identified players as candidates.

15. 2. The gaming table system according to claim 1, wherein the second matching means identifies the player ID of a player appearing in an image captured by the second camera based also on the combinations determined by the combination determination means for multiple players seated at other gaming tables in the casino.

16. 2. The gaming table system of claim 1, wherein when the second matching means cannot match the identified player ID with the player appearing in the image taken by the second camera, the second matching means matches the identified player ID with the player appearing in the other image using other images taken by the second camera before and after the image.

17. The game device further includes a database storing the player IDs of registered players and reference images in association with each other; 2. The gaming table system according to claim 1, wherein the reference image acquisition means refers to the database to identify the reference image of the player captured in the image taken by the second camera.

18. a display means for displaying the reference image specified by the reference image acquisition means; an input receiving means for receiving an input of approval or rejection from the dealer with respect to the reference image displayed on the display means; Furthermore, 18. The gaming table system of claim 17, wherein the reference image acquisition means acquires the reference image of the player who has received the rejection input from the input acceptance means by photographing the player, and associates the reference image with the player ID of the player and stores the reference image in the database.

19. The gaming table system of claim 1, wherein the chip recognition means comprises a first chip recognition means for recognizing the position of chips from an image generated by photographing the chips on the gaming table, and a second chip recognition means for recognizing the type and number of chips from an image generated by photographing the chips on the gaming table.

20. the chip recognition means recognizes the position, type, and number of the chip by reading the RFID tag embedded in the chip; the body recognition means recognizes the position of a player playing a game at the gaming table based on the image; 2. The gaming table system according to claim 1, wherein the first association means associates the player with the chip based on the position of the player recognized by the body recognition means and the position of the chip recognized by the chip recognition means.

Images