System for learning recognition of game coins

Abstract

This invention provides a game coin recognition learning system capable of accurately identifying game coins wagered by a player. The game coin recognition learning system (10) comprises: a game recording device (11) that uses a camera (212) to record the state of game coins (W) stacked on a game table (4) as an image; a game coin judging device (12) having an artificial intelligence device (12a) that performs image analysis on the recorded image of the state of the game coins (W) to determine the number and type of game coins (W) wagered by the player (C); and a teacher device (13) that, if it is suspected that the judgment result of the game coin judging device (12) is incorrect, inputs the image used in the judgment of the game coin judging device (12) and the correct number and type of game coins W for the incorrect judgment as teacher data into the artificial intelligence device (12a) and enables the artificial intelligence device (12a) to learn.

Description

[0001] This invention is a divisional application of the invention application with application number 201810063498.4 and invention title "Game Coin Recognition and Learning System". Technical Field

[0002] This invention relates to a learning system for recognizing game coins. Background Technology

[0003] In games such as [specific games], guests (players) stack multiple game tokens on a table to place bets. Therefore, accurate identification of the stacked game tokens is necessary. Furthermore, an example of game tokens used in a game is disclosed in International Publication No. 2008 / 120749. Summary of the Invention

[0004] The purpose of this invention is to provide a game currency recognition and learning system that can accurately identify the game currency wagered by players.

[0005] One aspect of the present invention is a game coin recognition learning system for use in an amusement park with game tables. The game coin recognition learning system comprises: a game recording device that uses a camera to record the state of game coins stacked on the game table as an image; a game coin judging device including an artificial intelligence device that performs image analysis on the recorded image of the game coin state to determine the number and type of game coins wagered by the player; and a teacher device that, if it is suspected that the judgment result of the game coin judging device is erroneous, inputs the image used in the judgment of the game coin judging device, along with the correct number and type of game coins for the erroneous judgment, as teacher data into the artificial intelligence device, and enables the artificial intelligence device to learn.

[0006] In this manner, when the game coin counting device is suspected of making an error, the teacher device inputs the images used in the counting process, along with the correct number and type of coins for the incorrect count, into the artificial intelligence device as teacher data. This allows the AI ​​device to learn, enabling it to effectively learn from image patterns where the game coin counting device has relatively low accuracy, thus improving the device's accuracy by focusing on those patterns. Through repeated teacher learning, the game coin counting device can accurately identify the coins bet by the player, regardless of how the coins are stacked.

[0007] In one aspect of the game coin recognition learning system of the present invention, if the teacher device determines that the determination result of the game coin determination device is correct, it may further input the image used in the determination of the game coin determination device, as well as the number and type of game coins in the determination result, as teacher data into the artificial intelligence device, so that the artificial intelligence device can learn.

[0008] In this way, not only can the accuracy of the game coin determination device be improved for image patterns with relatively low determination accuracy, but also for image patterns with relatively high determination accuracy. In this way, the game coin determination device can further improve the accuracy of identifying the game coins bet by the player.

[0009] One aspect of the present invention is a game coin recognition learning system for use in an amusement park with game tables. The game coin recognition learning system comprises: a game recording device that uses a camera to record the state of game coins stacked on the game table as an image; and a game coin judging device, which includes an artificial intelligence device. The artificial intelligence device performs image analysis on the recorded images of the game coin state to determine the number and type of game coins bet by the player. If the artificial intelligence device suspects that the judgment result of the game coin judging device is incorrect, it inputs the image used in the judgment of the game coin judging device and the correct number or type of game coins for the incorrect judgment as teacher data through a teacher device for learning.

[0010] In this way, when the game coin counting device is suspected of making an error, the AI ​​device learns from the images used in the counting process, along with the correct number and type of coins for the incorrect counts, by inputting them into a teacher device. This allows the AI ​​device to effectively learn from image patterns where the counting device's accuracy is relatively low, and to focus on improving those patterns. Through repeated processing, regardless of the way the coins are stacked, the game coin counting device can accurately identify the coins bet by the player.

[0011] In one aspect of the game coin recognition learning system of the present invention, a control device may also be included. This control device determines whether the determination result of the game coin determination device is correct. The game coin determination device, during a game played at the game table, can determine the type and number of game coins in the game coin tray of the game table, as well as the position, type, and number of game coins wagered by each player, based on images recorded in the game recording device. When all the game coins wagered by each player are recovered, the control device knows the actual total amount of game coins in the game coin tray. Based on the determination result of the game coin determination device, it calculates the expected total amount of game coins in the game coin tray and compares the expected total amount with the actual total amount. If there is a difference between the expected total and the actual total, it is determined that the determination result of the game coin determination device is suspected to be incorrect. The expected total amount of game coins in the game coin tray is obtained by adding the total amount of game coins in the game coin tray before each game's settlement to the increase in the game coin tray amount calculated based on the type and number of game coins wagered by the losing players.

[0012] In this way, the control device can automatically determine whether the determination result of the game coin determination device is suspected of being erroneous.

[0013] One aspect of the present invention is a game coin recognition learning system for use in an amusement park with game tables. The game coin recognition learning system is characterized by comprising: a game recording device that uses a camera to record the state of game coins stacked on the game table as an image; a game coin judging device, which includes an artificial intelligence device that performs image analysis on the recorded image of the game coin state to determine the number and type of game coins wagered by the player; and a teacher device that, if the judgment result of the game coin judging device is determined to be correct, inputs the image used in the judgment of the game coin judging device, as well as the number and type of game coins in the judgment result, as teacher data into the artificial intelligence device, and enables the artificial intelligence device to learn.

[0014] One aspect of the present invention is a game coin recognition learning system for use in an amusement park with game tables. The game coin recognition learning system is characterized by comprising: a game recording device that uses a camera to record the state of game coins stacked on the game table as an image; and a game coin judging device, which includes an artificial intelligence device. The artificial intelligence device performs image analysis on the recorded image of the game coin state to determine the number and type of game coins wagered by the player. If the artificial intelligence device determines that the judgment result of the game coin judging device is correct, it inputs the image used in the judgment of the game coin judging device and the number or type of game coins in the judgment result as teacher data through a teacher device for learning.

[0015] In one aspect of the game coin recognition learning system of the present invention, a control device may also be included. This control device determines whether the judgment result of the game coin judgment device is correct. The game coin judgment device, during a game played at the game table, can determine the type and number of game coins in the game coin tray of the game table, as well as the position, type, and number of game coins wagered by each player, based on images recorded in the game recording device. When all the game coins wagered by each player are recovered, the control device knows the actual total amount of game coins in the game coin tray. Based on the judgment result of the game coin judgment device, it calculates the expected total amount of game coins in the game coin tray and compares the expected total amount with the actual total amount. When the expected total amount matches the actual total amount, the judgment result of the game coin judgment device is deemed correct. The expected total amount of game coins in the game coin tray is obtained by adding the increase in the game coin tray amount calculated based on the type and number of game coins wagered by the losing players to the total amount of game coins in the game coin tray before each game's settlement.

[0016] In one embodiment of the game coin identification and learning system of the present invention, the control device may use the RFID tag installed on the game coin as a reference to determine the actual total amount of game coins in the game coin tray.

[0017] In this way, the control device can use RFID to automatically determine the actual total amount of game coins in the coin tray, which improves measurement accuracy compared to visual estimation by staff.

[0018] In one aspect of the game coin recognition learning system of the present invention, the control device may have an artificial intelligence device for correct solution determination that is different from the artificial intelligence device of the game coin determination device. This artificial intelligence device for correct solution determination determines the actual total amount of game coins in the game coin tray based on the image recorded in the game recording device.

[0019] In this way, the control device can use an artificial intelligence device for forward comprehension to automatically determine the actual total amount of game coins in the game coin tray, which can improve measurement accuracy compared to visual estimation by staff.

[0020] In one aspect of the game coin recognition learning system of the present invention, the game recording device may assign an index or time to the image obtained from the camera, or assign a tag to determine the recycling or payment scenario of the game coin, so that the game coin determination device can subsequently analyze the game record.

[0021] In this way, the game coin determination device can easily determine the state of the game coin that should be analyzed by using the index, time, and tag assigned to the image, based on the content recorded by the game recording device, thus shortening the time required for determination.

[0022] In one aspect of the game coin recognition learning system of the present invention, even when multiple game coins placed on the game table are hidden due to falling into the blind spot of the camera, resulting in some or all of the game coins being hidden, the game coin determination device can still determine the type, number, and location of the game coins being bet.

[0023] In this way, especially when multiple game tokens placed on the game table are hidden due to falling into the blind spot of the staff, the accuracy of the measurement can be improved by having the game token detection device determine the game tokens bet by the player, compared with the staff's visual estimation.

[0024] One aspect of the present invention provides a game coin recognition learning method for use in an amusement park with game tables. The game coin recognition learning method includes: a game recording step, in which the state of game coins stacked on the game table is recorded as an image using a camera; a game coin judgment step, performed by an artificial intelligence device, which analyzes the recorded image of the game coin state to determine the number and type of game coins wagered by the player; and a teacher step, in which, if it is suspected that the judgment result in the game coin judgment step is incorrect, the image used in the judgment of the game coin judgment step, along with the correct number and type of game coins for the incorrect judgment, are input into the artificial intelligence device as teacher data, and the artificial intelligence device learns from this data.

[0025] In this way, if an error is suspected in the coin-counting step, the image used in the coin-counting step, along with the correct number and type of coins for the incorrect coin-counting step, is input into the AI ​​device as teacher data during the teacher step. The AI ​​device then learns from this data, enabling it to effectively learn image patterns with relatively low accuracy in the coin-counting step and improve accuracy by focusing on those patterns. Through repeated iterations, the AI ​​can accurately identify the coins bet by the player regardless of how they are stacked in the coin-counting step.

[0026] One aspect of the present invention provides a game coin recognition learning method for use in an amusement park with game tables. The method is characterized by comprising: a game recording step, in which the state of game coins stacked on the game table is recorded as an image using a camera; a game coin determination step, performed by an artificial intelligence device, which analyzes the recorded image of the game coin state to determine the number and type of game coins wagered by the player; and a teacher step, in which, if the determination result in the game coin determination step is correct, the image used in the determination step and the number and type of game coins determined in the determination result are input into the artificial intelligence device as teacher data, and the artificial intelligence device learns from this data. Attached Figure Description

[0027] Figure 1 This is a diagram schematically illustrating an amusement park with a game coin recognition learning system according to the first embodiment.

[0028] Figure 2 It is a diagram used to illustrate the process of playing a game.

[0029] Figure 3This is a block diagram showing the general structure of the game coin recognition learning system according to the first embodiment.

[0030] Figure 4 This is a flowchart illustrating the learning method for recognizing game coins.

[0031] Figure 5 This is a flowchart illustrating a variation of a learning method for recognizing game coins.

[0032] Figure 6 This is a flowchart illustrating another variation of the learning method for recognizing game coins. Figure 7 This is a diagram schematically illustrating an amusement park with a game coin recognition learning system according to the second embodiment.

[0033] Figure 8 This is a block diagram illustrating the general structure of the game coin recognition learning system according to the second embodiment. Detailed Implementation

[0034] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. Furthermore, in each drawing, components having equivalent functions are given the same reference numerals, and detailed descriptions of components with the same reference numerals are not repeated.

[0035] First, a game played in an amusement park with game table 4 will be described. In this embodiment, an example of a game table with game table 4 being described is given, but the present invention can also be applied to other amusement parks or other games.

[0036] Figure 1 This is a schematic diagram of an amusement park equipped with a game coin recognition learning system 10 according to the first embodiment. (See diagram below.) Figure 1 As shown, the amusement park is equipped with a roughly semi-circular game table 4 and multiple chairs 201 arranged along the arc side of the game table 4, facing the dealer D. The number of chairs 201 is arbitrary. Figure 1 In the example shown, six chairs 201 are arranged. Additionally, a betting area BA is provided on the game table 4 corresponding to each chair 201. That is, in the illustrated example, the six betting areas BA are arranged in an arc shape.

[0037] like Figure 1 As shown, guests (players) C are seated in chairs 201. Guests (players) C place bets (hereinafter referred to as "placements") on the outcome of a game by stacking game coins W in the placement area BA set in front of the chair 201 they are sitting in. The outcome of a game includes which side, P (PLAYER) or B (BANKER), wins or a tie (TIE).

[0038] The game tokens W placed can be of only one type or multiple types. Alternatively, the number of game tokens W placed can be arbitrarily determined by the guest (player) C. The game token recognition learning system 10 of this embodiment can recognize the number and type of game tokens W in the stacked configuration.

[0039] To ensure that player C finishes placing cards, dealer D performs the following actions: starts a timer and announces "No More Bet," moves his hand horizontally, etc. Then, dealer D draws cards one by one from the card dealing device S onto the game table 4. (Example...) Figure 2 As shown, the first card is dealt to P (PLAYER), the second card is dealt to B (BANKER), the third card is dealt to P (PLAYER), and the fourth card is dealt to B (BANKER) (hereinafter, the drawing of the first to fourth cards is referred to as "dealing").

[0040] Furthermore, all cards are drawn face down from the card dealing device S. Therefore, neither the dealer D nor the player C can know the rank (number) or suit (hearts, diamonds, spades, clubs) of the drawn cards.

[0041] After the fourth card is drawn, the player (player) C who placed the card in area P (PLAYER) (in the case of multiple players placing cards in area P, it is the player C who placed the highest amount; in the case of no players placing cards in area P, it is the dealer D) turns the first and third cards face up. Similarly, the player (player) C who placed the card in area B (BANKER) (in the case of multiple players placing cards in area B, it is the player C who placed the highest amount; in the case of no players placing cards in area B, it is the dealer D) turns the second and fourth cards face up (generally, this act of turning face-up cards face up is called "squeeze").

[0042] Then, based on the rank (number) of the first to fourth cards and the detailed rules of the game, the dealer D draws the fifth card, then the sixth card, and deals these cards to the respective P (Player) or B (Banker). Similarly, the guest (player) C placed in the P area (Player) reveals the cards dealt to the P (Player), and the guest (player) placed in the B area (Banker) reveals the cards dealt to the B (Banker).

[0043] The most enjoyable part for guest (player) C is the time between the first to fourth cards being drawn and the fifth and sixth cards being revealed to determine the winner.

[0044] Furthermore, sometimes the outcome is decided after the first four cards, depending on the card ranking (number), and sometimes it takes until the fifth or even the sixth card. The dealer D performs the following tasks: based on the ranking (number) of the cards after they are dealt, he / she ascertains the outcome and presses the outcome display button on the card dealing device S, and displays the outcome on a monitor to inform the guest (player) C of the outcome.

[0045] Additionally, the game's outcome is determined simultaneously using the win / loss determination unit of the card dealing device S. An error occurs if, despite a win / loss being decided, the dealer D does not display the result but still draws cards. The card dealing device S detects this error and outputs an error signal. Finally, while displaying the win / loss result, the dealer D settles the bets of guest (player) C, paying out to the winning guest (player) C and recovering the bets of the losing guest (player) C. After settlement, the display of the win / loss result ends, and the next game begins.

[0046] Furthermore, the aforementioned game process is widely conducted in ordinary casinos. The card dealing device S is an existing card dealing device, which employs a structure where the dealer D manually draws out the cards. It can also read the drawn cards and has a result display button and a result display unit, enabling it to determine the winner and display the result. On the floor of an ordinary casino, each of the multiple game tables 4 is equipped with a card dealing device S and a monitor. The cards used are supplied to each game table 4 or a cabinet beneath each game table 4 in bundles, sets, or even boxes.

[0047] The game coin recognition and learning system 10 of this embodiment is a system for recognizing and learning game coins W stacked and arranged in the placement area BA by the guest (player) C, and more specifically, it is a system for recognizing and learning the number and / or type of game coins W.

[0048] like Figure 1As shown, in this embodiment, a surveillance camera 212 is provided on the outside of the game table 4 to capture the state of the game coins W stacked and arranged in the placement area BA. In addition, each game coin W is equipped with RFID (Radio Frequency Identification), and an RFID reader 22 is provided on the game coin tray 23 managed by the dealer D to read the RFID of the game coins W in the game coin tray 23.

[0049] In this embodiment, the game coin recognition and learning system 10 is connected to the surveillance camera 212 and the RFID reader 22 in a communicable manner.

[0050] Figure 3 This is a block diagram showing the general configuration of the game coin recognition and learning system 10 of this embodiment.

[0051] like Figure 3 As shown, the game coin recognition learning system 10 includes a game recording device 11, a game coin judging device 12, a teacher device 13, and a control device 14. Furthermore, at least a portion of the game coin recognition learning system 10 is implemented by a computer.

[0052] The game recording device 11 has a solid-state data storage device such as a hard disk. The game recording device 11 records the state of the game coins W stacked on the game table 4 as an image captured by the camera 212. In addition, the image can be a moving image or a continuous still image.

[0053] The game recording device 11 can also record by indexing or assigning a time to the image obtained from the camera 212, or by assigning a tag to determine the recycling or payment scenario of the game currency W, so that the game record can be analyzed later by the game currency determination device 12 described later.

[0054] The game coin determination device 12 includes an artificial intelligence device 12a that uses image recognition technology, such as deep learning, to analyze the image of the state of the game coin W recorded by the game recording device 11 to determine the number and type of game coins W bet by the customer (player) C. The game coin determination device 12 can also further determine the position of the game coin W bet by the customer (player) C on the placement area BA.

[0055] The game coin determination device 12 can also perform image analysis on the image of the state of the game coins W recorded in the game recording device 11 to determine the number and type of game coins W in the game coin tray 23 before each game is settled.

[0056] like Figure 3As shown, the game coin determining device 12 outputs the determining result to the output device 15. The output device 15 can output the determining result of the game coin determining device 12 as text information to the monitor on the game table 4, or as audio information to the dealer D's headphones, etc.

[0057] Control device 14 is used to determine whether the determination result of game coin determination device 12 is correct. When the game coins W (used game coins) wagered by the losing guest (player) C are all recovered, control device 14 knows the actual total amount V0 of game coins W in game coin tray 23.

[0058] In this embodiment, the control device 14 obtains the RFID information of the game coins W in the game coin tray 23 from the RFID reading device 22, and determines the type and number of game coins W in the game coin tray 23 based on the obtained RFID information, and knows the actual total amount V0 of game coins W in the game coin tray 23.

[0059] Additionally, the control device 14 obtains the determination result from the game coin determination device 12. Based on the obtained determination result, it calculates the total amount V1 according to the type and number of game coins W in the game coin tray 23 before each game's settlement, and calculates the total amount of game coins W bet by the losing player C (i.e., the increase in the game coin tray 23 for that game) V2 according to the position, type, and number of game coins W bet by each player C. Then, the control device 14 adds the increase in the game coin tray 23 for that game to the total amount V1 of game coins W in the game coin tray 23 before each game's settlement, thereby calculating the proper total amount V3 of game coins in the game coin tray 23 (=V1+V2).

[0060] The control device 14 compares the expected total amount V3 of game coins W in the game coin tray 23 with the actual total amount V0 of game coins W in the game coin tray 23. When there is a difference between the expected total amount V3 and the actual total amount V0 (V3 ≠ V0), it determines that the determination result of the game coin determination device 12 is suspected to be incorrect. On the other hand, when the expected total amount V3 and the actual total amount V0 are the same (V3 = V0), the control device 14 determines that the determination result of the game coin determination device 12 is correct.

[0061] When the game coins W are recovered from the losing player C, the winning player C is paid out game coins W. The control device 14 calculates the total amount of game coins W wagered by the winning player C, and the corresponding payable amount V4, based on the position, type, and number of game coins W wagered by each player C. The control device 14 monitors the actual total amount of game coins W in the game coin tray 23 after the payout, determines whether this actual total amount matches the payable amount V4, and displays a light indicating whether they match based on the determination result.

[0062] Control device 14 compares the expected total amount V5 (=V1+V2-V4) of game coins W in game coin tray 23 with the actual total amount of game coins W in game coin tray 23 after the addition of recycled game coins and the reduction of paid game coins. If there is a difference, it determines that the judgment result of game coin judgment device 12 is suspected to be incorrect. If the expected total amount V5 is consistent with the actual total amount, control device 14 determines that the judgment result of game coin judgment device 12 is correct.

[0063] To determine whether the lights are consistent, for example, the lights can be turned on in a way that turns green if they are consistent and red if they are inconsistent.

[0064] The teacher device 13 obtains a judgment from the control device 14 regarding the correctness of the judgment result from the game coin judging device 12. If the control device 14 suspects an error in the judgment result of the game coin judging device 12, the teacher device 13 inputs the image used in the judgment of the game coin judging device 12 (including the suspected error) along with the correct number and type of game coins W for the error as teacher data into the artificial intelligence device 12a of the game coin judging device 12, and enables the artificial intelligence device 12a to learn. Furthermore, the correct number and type of game coins for the error are actually verified by a human and taught to the teacher device 13. That is, regarding the correct number and type of game coins for the error, the teacher device 13 learns by having a human demonstrate the image at the time of the error and the correct number of coins at that time.

[0065] The teacher device 13 can also input the image used in the (correct) judgment of the game coin judgment device 12 and the number and type of game coins W in the judgment result of the game coin judgment device 12 (i.e. the correct number and type of game coins W) as teacher data into the artificial intelligence device 12a of the game coin judgment device 12, and enable the artificial intelligence device 12a to learn.

[0066] The teacher device 13 repeatedly inputs the aforementioned teacher data into the artificial intelligence device 12a of the game coin determination device 12, enabling the AI ​​device 12a to learn such teacher actions. This improves the accuracy of the game coin determination device 12 in determining the game coin W. The AI ​​device 12a of the game coin determination device 12 performs image analysis on the image of the state of the game coin W to determine the game coin W. Therefore, even if multiple game coins W placed on the game table 4 are partially or completely hidden due to falling into the blind spot of the camera 212, the type, number, and position of the bet game coin W can still be determined by repeatedly learning from such incomplete images.

[0067] Next, refer to Figure 4 The operation (game coin recognition and learning method) of the game coin recognition and learning system 10 of this embodiment will be explained.

[0068] like Figure 4 As shown, firstly, when the guest (player) C stacks game coins W in the placement area BA of the game table 4, the camera 212 captures the state of the stacked game coins W as an image, and the game recording device 11 records the image (step S31).

[0069] Next, the game coin determination device 12 performs image analysis on the image recorded in the game recording device 11 to determine the number and type of game coins W wagered by the customer (player) C (step S32). Furthermore, the image analyzed by the game coin determination device 12 can also be an image selected based on an index, time, or a tag used by the game recording device 11 to determine the recycling or payment scenario of the game coin W.

[0070] In step S32, the image of the state of the game coins W recorded in the game recording device 11 can also be analyzed by the game coin determination device 12. This analysis can determine not only the number and type of game coins W bet by the customer (player) C, but also the position of the game coins W bet by the customer (player) C on the placement area BA. It can also determine the number and type of game coins W in the game coin tray 23 before each game is settled.

[0071] The result of the game coin determination device 12 is output to the output device 15. The result of the game coin determination device 12 can be output as text information to the monitor on the game table 4 via the output device 15, or as audio information to the dealer D's headphones, etc.

[0072] The determination result of the game coin determination device 12 is also sent to the control device 14, which determines whether the determination result of the game coin determination device 12 is correct (step S33).

[0073] If the control device 14 determines that the judgment result of the game coin judgment device 12 is suspected to be incorrect (step S34: no), the image used in the judgment of the game coin judgment device 12 (including the suspected error) and the number and type of correct game coins W for the error are input from the teacher device 13 to the artificial intelligence device 12a of the game coin judgment device 12 as teacher data, and the artificial intelligence device 12a learns (step S36).

[0074] On the other hand, if the control device 14 determines that the determination result of the game coin determination device 12 is correct (step S34: yes), the operation of the game coin recognition learning system 10 in the game ends.

[0075] As described above, according to this embodiment, when it is suspected that the judgment result of the coin judging device 12 is erroneous, the teacher device 13 inputs the image used in the judgment of the coin judging device 12, as well as the correct number and type of coins W for the erroneous judgment, as teacher data into the artificial intelligence device 12a, and enables the artificial intelligence device 12a to learn. Therefore, the artificial intelligence device 12a can effectively learn from image patterns where the judgment accuracy of the coin judging device 12 is relatively low, and can focus on improving the judgment accuracy of the coin judging device 12 by targeting these image patterns. By repeatedly performing this teacher learning, regardless of how the coins W are stacked, the coin judging device 12 can accurately identify the coins W bet by the player C.

[0076] Furthermore, according to this embodiment, when all the game coins wagered by each player C are recovered, the control device 14 determines the actual total amount V0 of the game coins W in the game coin tray 23. Based on the determination result of the game coin determination device 12, it calculates the proper total amount V3 of the game coins W in the game coin tray 23 (wherein, the proper total amount of the game coins W in the game coin tray 23 is the total amount V1 of the game coins W in the game coin tray 23 before the settlement of each game, plus the type and number of game coins W wagered by the losing player C). The increase V2 of the game coin tray 23 in the game is calculated. V3 = V1 + V2). The total amount of game coins W in the game coin tray 23, V3, is compared with the actual total amount of game coins W in the game coin tray 23, V0. When there is a difference between the total amount V3 and the actual total amount V0 (V3 ≠ V0), it is determined that the judgment result of the game coin judgment device 12 is suspected to be wrong. In this way, the control device 14 can automatically determine whether the judgment result of the game coin judgment device 12 is suspected to be wrong.

[0077] Furthermore, according to this embodiment, the control device 14 uses the RFID installed on the game coin W as a reference to determine the actual total amount V0 of the game coins W in the game coin tray 23. Therefore, the control device 14 can use RFID to automatically determine the actual total amount V0 of the game coins W in the game coin tray 23, which can improve the measurement accuracy compared to visual inspection by the staff.

[0078] Furthermore, according to this embodiment, the game recording device 11 records images acquired from the camera 212 by assigning an index or time, or by assigning a tag for determining the recycling or payment scenario of game coins. Therefore, by utilizing the index, time, and tag assigned to the image, the game coin determination device 12 can easily determine the image of the state of the game coin W that should be analyzed based on the recorded content of the game recording device 11, thereby shortening the time required for determination.

[0079] Furthermore, according to this embodiment, even when multiple game tokens W placed on the game table 4 are hidden due to falling into the blind spot of the camera 212, resulting in some or all of the game tokens W being hidden, the game token determination device 12 can still determine the type, number, and position of the game tokens W bet. Therefore, especially in the case where multiple game tokens W placed on the game table 4 are hidden due to falling into the blind spot of the staff, by having the game token determination device 12 determine the game tokens W bet by the player C, the measurement accuracy can be improved compared to the staff's visual inspection.

[0080] Furthermore, various modifications can be made to the above-described embodiments. Hereinafter, an example of a modification will be described with reference to the accompanying drawings. In the following description and the accompanying drawings used in the description, parts that can be constructed in the same way as the parts corresponding to those in the above-described embodiments are used with the same reference numerals, and repeated descriptions are omitted.

[0081] Figure 5 This is a flowchart illustrating a variation of a learning method for recognizing game coins.

[0082] exist Figure 5 In the example shown, if the control device 14 determines that the judgment result of the game coin judgment device 12 is suspected to be incorrect (step S34: no), the image used in the (suspected error) judgment of the game coin judgment device 12, as well as the number and type of correct game coins W for the error, are input from the teacher device 13 to the artificial intelligence device 12a of the game coin judgment device 12 as teacher data, and the artificial intelligence device 12a is further made to learn (step S36).

[0083] On the other hand, if the control device 14 determines that the determination result of the game coin determination device 12 is correct (step S34: Yes), the image used in the (correct) determination of the game coin determination device 12 and the number and type of game coins W in the determination result of the game coin determination device 12 (i.e. the correct number and type of game coins W) are used as teacher data and further input from the teacher device 13 to the artificial intelligence device 12a of the game coin determination device 12, and the artificial intelligence device 12a is further made to learn (step S35).

[0084] In this way, not only can the accuracy of the game coin determination device 12 be improved for image patterns with relatively low determination accuracy, but also for image patterns with relatively high determination accuracy. In this way, the game coin determination device 12 can further identify the game coins W bet by player C with higher accuracy.

[0085] Figure 6 This is a flowchart illustrating another variation of the learning method for recognizing game coins.

[0086] exist Figure 6 In the example shown, if the control device 14 determines that the determination result of the game coin determination device 12 is correct (step S34: Yes), the image used in the (correct) determination of the game coin determination device 12 and the number and type of game coins W in the determination result of the game coin determination device 12 (i.e. the correct number and type of game coins W) are further input from the teacher device 13 to the artificial intelligence device 12a of the game coin determination device 12 as teacher data, and the artificial intelligence device 12a is further made to learn (step S35).

[0087] On the other hand, if the control device 14 determines that the determination result of the game coin determination device 12 is suspected to be incorrect (step S34: no), the operation of the game coin recognition learning system 10 in the game is terminated.

[0088] In this manner, when the game coin determination device 12 determines the result to be correct, the teacher device 13 inputs the image used in the determination by the game coin determination device 12, as well as the number and type of (correct) game coins W in the determination result, as teacher data into the artificial intelligence device 12a, enabling the artificial intelligence device 12a to learn. Therefore, the artificial intelligence device 12a can effectively learn from image patterns with relatively high determination accuracy of the game coin determination device 12, and can focus on improving the determination accuracy of the game coin determination device 12 by targeting these image patterns. Through repeated teacher learning in this way, the game coin determination device 12 can accurately identify the game coins W bet by player C.

[0089] Figure 7 This is a schematic diagram of an amusement park having a game coin recognition learning system 100 according to the second embodiment. Figure 8 This is a block diagram showing the schematic configuration of the game coin recognition and learning system 100 according to the second embodiment.

[0090] like Figure 7 As shown, in the second embodiment, in addition to the surveillance camera 212 used to capture the state of the game coins W stacked in the placement area BA, a surveillance camera 24 for the game coin tray is also provided on the outside of the game table 4 to capture the state of the game coins W in the game coin tray 23 managed by the dealer D.

[0091] The game coin recognition and learning system 100 of the second embodiment is communicatively connected to the surveillance camera 212 and the game coin tray surveillance camera 24, respectively.

[0092] like Figure 8 As shown, the game recording device 11 records the state of the game coin W in the game coin tray 23 as an image captured by the camera 24 on the game coin tray. Furthermore, the image can be a moving image or a continuous static image.

[0093] The control device 14 has an artificial intelligence device 14a (for correct solution determination) that is different from the artificial intelligence device 12a of the game coin determination device 12. It performs image analysis on the image of the state of the game coins W in the game coin tray 23 recorded by the game recording device 11, determines the number and type of game coins W in the game coin tray 23, and grasps the actual total amount V0 of the game coins W in the game coin tray 23. The artificial intelligence device 14a uses image recognition technology such as deep learning.

[0094] According to the second embodiment, the control device 14 can use the artificial intelligence device 14a for correct solution determination to automatically determine the actual total amount of game coins W in the game coin tray 23, which can improve the measurement accuracy compared with the visual estimation of the staff.

[0095] Furthermore, there is a technique commonly referred to as teacherless data learning, but this teacherless data learning is a technique that teaches artificial intelligence to determine whether the result is correct or incorrect; this is also referred to as teacher-based data learning in this invention. The above embodiments are described with the aim of enabling those skilled in the art to implement the invention. As various modifications of the above embodiments can obviously be implemented by those skilled in the art, the technical concept of the invention can also be applied to other embodiments. Therefore, the invention is not limited to the described embodiments, but should be considered within the broadest scope of the technical concept defined by the scope of the claims.

Claims

1. A game coin recognition learning system for use in an amusement park with game tables, characterized in that it has: A game recording device that uses a camera to capture images of the game coins stacked on the game table and records them. The game coin determination device includes an artificial intelligence device that has been trained in advance using teacher data. The artificial intelligence device performs image analysis on the recorded images of the state of the game coins to determine the number and type of game coins placed by the player on the game table. as well as The teacher device, when it suspects that the game coin determination device's result is erroneous, inputs the image used in the game coin determination device's determination, along with the correct number and type of game coins for the erroneous determination, as teacher data into the artificial intelligence device, enabling the artificial intelligence device to perform additional learning. Even when multiple game tokens placed on the game table fall into the blind spot of the camera, causing some or all of the game tokens to be hidden, the game token detection device can still determine the type, number, and location of the game tokens.

2. The game coin recognition and learning system as described in claim 1, characterized in that, If the teacher device determines that the determination result of the game coin determination device is correct, it further inputs the image used in the determination of the game coin determination device, as well as the number and type of game coins in the determination result, into the artificial intelligence device as teacher data, so that the artificial intelligence device can learn.

3. A game coin recognition learning system for use in an amusement park with game tables, characterized in that it has: A game recording device that uses a camera to capture images of the game coins stacked on the game table and records them; and The game coin determination device includes an artificial intelligence device that has been pre-learned from teacher data. This AI device performs image analysis on recorded images of the game coin's state to determine the number and type of game coins placed by the player on the game table. If the AI ​​device suspects an error in the game coin authentication device's result, it inputs the image used in the game coin authentication device's authentication process, along with the correct number or type of game coins for the error, into the teacher device as teacher data for further learning. Even when multiple game tokens placed on the game table fall into the blind spot of the camera, causing some or all of the game tokens to be hidden, the game token identification device can still determine the type, number, and location of the game tokens being bet.

4. The game coin recognition and learning system as described in claim 1 or 3, characterized in that, It also includes a control device that determines whether the determination result of the game coin determination device is correct. The game coin determining device, during a game played at the arcade table, can determine the type and quantity of game coins in the game coin tray of the arcade table, as well as the position, type, and quantity of game coins wagered by each player, based on images recorded in the game recording device. When all the game coins wagered by each player are recovered, the control device determines the actual total amount of game coins in the game coin tray. Based on the determination result of the game coin determination device, it calculates the expected total amount of game coins in the game coin tray and compares the expected total amount of game coins in the game coin tray with the actual total amount of game coins in the game coin tray. If there is a difference between the expected total amount and the actual total amount, it is determined that the determination result of the game coin determination device is suspected to be incorrect. The expected total amount of game coins in the game coin tray is obtained by adding the total amount of game coins in the game coin tray before the settlement of each game to the increase in the game coin tray for that game, calculated based on the type and number of game coins wagered by the losing players.

5. The game coin recognition and learning system as described in claim 4, characterized in that, The control device uses the RFID tag installed on the game coins as a reference to determine the actual total amount of game coins in the game coin tray.

6. The game coin recognition and learning system as described in claim 4, characterized in that, The control device has an AI device for determining the correct answer, which is different from the AI ​​device of the game coin determination device. This AI device for determining the correct answer determines the actual total amount of game coins in the game coin tray based on the image recorded in the game recording device.

7. The game coin recognition and learning system as described in any one of claims 4 to 6, characterized in that, The game recording device records images acquired from the camera by assigning an index or time, or by assigning a context for determining the redemption or payment of game coins, so that the game coin determination device can analyze the game records later.

8. A method for recognizing game coins, used in an amusement park with game tables, characterized in that the method includes: The game recording process involves using a camera to capture images of the game coins stacked on the game table and recording the images. The game coin determination step is performed by an artificial intelligence device that has been trained on teacher data in advance. The artificial intelligence device performs image analysis on the recorded images of the state of the game coins to determine the number and type of game coins placed by the player on the game table. as well as In the teacher step, if it is determined that the judgment result in the game coin judgment step is suspected to be incorrect, the image used in the judgment of the game coin judgment step, as well as the correct number and type of game coins for the error, are input into the artificial intelligence device as teacher data, and the artificial intelligence device performs additional learning. Even when multiple game tokens placed on the game table fall into the blind spot of the camera, causing some or all of the game tokens to be hidden, the game token identification device can still determine the type, number, and location of the game tokens being bet.

9. A game coin recognition learning system for use in an amusement park with game tables, characterized in that it has: A game recording device that uses a camera to capture images of the game coins stacked on the game table and records them. The game coin determination device includes an artificial intelligence device that has been trained in advance using teacher data. The artificial intelligence device performs image analysis on the recorded image of the state of the game coins to determine the number and type of game coins placed by the player on the game table. as well as, The teacher device, upon determining that the game coin determination device's result is correct, inputs the image used in the game coin determination device's determination, as well as the number and type of game coins in the determination result, as the teacher data into the artificial intelligence device, enabling the artificial intelligence device to perform additional learning. Even when multiple game tokens placed on the game table fall into the blind spot of the camera, causing some or all of the game tokens to be hidden, the game token identification device can still determine the type, number, and location of the game tokens being bet.

10. A game coin recognition learning system for use in an amusement park with game tables, characterized in that it has: A game recording device that uses a camera to capture images of the game coins stacked on the game table and records them. The game coin determination device includes an artificial intelligence device that has been pre-learned from teacher data. This AI device performs image analysis on the recorded images of the game coin's state to determine the number and type of game coins placed by the player on the game table. If the AI ​​device determines that the game coin determination device's result is correct, it inputs the image used in the game coin determination device's determination and the number or type of game coins in the determination result into the teacher device as teacher data for further learning. Even when multiple game tokens placed on the game table fall into the blind spot of the camera, causing some or all of the game tokens to be hidden, the game token identification device can still determine the type, number, and location of the game tokens being bet.

11. The game coin recognition and learning system as described in claim 9 or 10, characterized in that, It also includes a control device that determines whether the determination result of the game coin determination device is correct. The game coin determining device, during a game played at the arcade table, can determine the type and quantity of game coins in the game coin tray of the arcade table, as well as the position, type, and quantity of game coins wagered by each player, based on images recorded in the game recording device. When all the game coins wagered by each player are recovered, the control device measures the actual total amount of game coins in the game coin tray. Based on the judgment result of the game coin judgment device, it calculates the expected total amount of game coins in the game coin tray and compares the expected total amount of game coins in the game coin tray with the actual total amount of game coins in the game coin tray. When the expected total amount and the actual total amount are consistent, the judgment result of the game coin judgment device is determined to be correct. The expected total amount of game coins in the game coin tray is obtained by adding the total amount of game coins in the game coin tray before settlement to the increase in the game coin tray for that game, calculated based on the type and number of game coins wagered by the losing players.

12. The game coin recognition and learning system as described in claim 11, characterized in that, The control device uses the RFID tag installed on the game coins as a reference to determine the actual total amount of game coins in the game coin tray.

13. The game coin recognition and learning system as described in claim 11, characterized in that, The control device has an AI device for determining the correct answer, which is different from the AI ​​device of the game coin determination device. This AI device for determining the correct answer determines the actual total amount of game coins in the game coin tray based on the image recorded in the game recording device.

14. The game coin recognition and learning system as described in any one of claims 11 to 13, characterized in that, The game recording device records images acquired from the camera by assigning an index or time, or a tag to determine the recycling or payment scenario of game coins, in a manner that allows the game coin determination device to analyze the game's records afterward.

15. A method for recognizing game coins, used in an amusement park with game tables, characterized in that the method includes: The game recording process involves using a camera to capture images of the game coins stacked on the game table and recording the images. The game coin determination step is performed by an artificial intelligence device that has been trained on teacher data in advance. The artificial intelligence device performs image analysis on the recorded image of the state of the game coins to determine the number and type of game coins placed by the player on the game table. as well as In the teacher step, if the determination result in the game coin determination step is correct, the image used in the determination of the game coin determination step, as well as the number and type of game coins in the determination result, are input into the artificial intelligence device as teacher data, and the artificial intelligence device performs additional learning. Even when multiple game tokens placed on the game table fall into the blind spot of the camera, causing some or all of the game tokens to be hidden, the game token identification device can still determine the type, number, and location of the game tokens being bet.

Images