Program, information processing method, and information processing device

Abstract

To provide a program, etc. capable of calculating winning points in a mah-jongg game in a general manner.SOLUTION: A program causes a computer to execute processing for acquiring a winning image capturing a winning hand composed of a plurality of mah-jongg tiles in a mah-jongg game, outputting a type of each mah-jongg tile included in the winning image by inputting the winning image to a deep layer learning model that has learned to output types of mah-jongg tiles included in the winning image when the winning image is input, and calculating winning points on the basis of the type of each mah-jongg tile that has been output.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present technology relates to a program, an information processing method, and an information processing device. [Background technology]

[0002] Conventionally, mahjong tables that calculate points for mahjong games have been proposed. For example, in the mahjong table described in Patent Document 1, the type data of mahjong tiles with built-in non-contact digital IC tags is read, and a computer calculates the points based on the type data. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Publication No. 2010-253015 Summary of the Invention [Problem to be solved by the invention]

[0004] However, the mahjong table in Patent Document 1 is only capable of reading tile type data from mahjong tiles with built-in IC tags, and is therefore lacking in versatility.

[0005] The present disclosure has been made in consideration of the above circumstances, and aims to provide a program or the like that can generically calculate the winning points in a mahjong game. [Means for solving the problem]

[0006] A program according to one embodiment of the present disclosure acquires a winning image obtained by photographing a winning hand consisting of multiple mahjong tiles in a mahjong game, and inputs the winning image into a deep learning model that has been trained to output the types of mahjong tiles contained in the winning image when a winning image is input, thereby outputting the types of each mahjong tile contained in the winning image, and causes a computer to execute a process of calculating a winning score based on the types of each mahjong tile output.

[0007] An information processing method according to one embodiment of the present disclosure obtains a winning image obtained by photographing a winning hand consisting of multiple mahjong tiles in a mahjong game, and inputs the winning image into a deep learning model that has been trained to output the types of mahjong tiles contained in the winning image when a winning image is input, thereby outputting the types of each mahjong tile contained in the winning image, and calculating a winning score based on the types of each mahjong tile output.

[0008] An information processing device according to one embodiment of the present disclosure acquires a winning image obtained by photographing a winning hand consisting of multiple mahjong tiles in a mahjong game, and outputs the types of mahjong tiles contained in the winning image by inputting the winning image into a deep learning model that has been trained to output the types of mahjong tiles contained in the winning image when a winning image is input, and is equipped with a control unit that outputs a winning score based on the types of each mahjong tile output. Effect of the Invention

[0009] In a program according to an embodiment of the present disclosure, it is possible to generally calculate the winning points in a mahjong game. [Brief description of the drawings]

[0010] [Figure 1] 1 is a block diagram showing an example of the configuration of an information processing device; [Diagram 2] An explanatory diagram showing a winning hand. [Diagram 3] An explanatory diagram explaining a tile detection model and a tile position discrimination module. [Figure 4] FIG. 13 is an explanatory diagram showing a shooting instruction screen. [Diagram 5] FIG. 13 is an explanatory diagram showing a score display screen. [Figure 6] FIG. 13 is an explanatory diagram showing a dragon input screen. [Figure 7] 11 is a flowchart illustrating an example of processing by the information processing device. [Figure 8]FIG. 11 is a block diagram showing an example of the configuration of an information processing device according to a second embodiment. [Figure 9] FIG. 1 is an explanatory diagram showing a mahjong table. [Figure 10] FIG. 11 is an explanatory diagram illustrating a tile detection model and a tile position discrimination module according to the second embodiment. [Figure 11] FIG. 11 is an explanatory diagram for explaining a tabletop information detection model. [Figure 12] FIG. 11 is an explanatory diagram showing a shooting instruction screen according to the second embodiment. [Figure 13] FIG. 11 is an explanatory diagram showing a score display screen according to the second embodiment. [Figure 14] 11 is a flowchart showing an example of processing by an information processing device according to the second embodiment. [Figure 15] FIG. 13 is an explanatory diagram showing a pre-climbing image. [Figure 16] FIG. 11 is an explanatory diagram showing a shooting instruction screen according to the third embodiment. [Figure 17] FIG. 13 is an explanatory diagram showing a listening count display screen. [Figure 18] An explanatory diagram showing the waiting tile display screen. [Figure 19] 11 is a flowchart showing an example of processing by an information processing device according to a third embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, the present invention will be described in detail with reference to the drawings showing embodiments thereof.

[0012] (Embodiment 1) 1 is a block diagram showing a configuration example of an information processing device 1. The information processing device 1 is an information processing device that processes, stores, and transmits / receives various information. The information processing device 1 is, for example, an information processing device such as a smartphone, a mobile phone, a wearable device such as an Apple Watch (registered trademark), a wearable camera, a tablet, or a personal computer. In the following description, the information processing device 1 is assumed to be a smartphone.

[0013] The information processing device 1 includes a control unit 11, a storage unit 12, an input unit 13, a display unit 14, and an image capturing unit 15. Each component is connected to a bus B.

[0014] The control unit 11 includes an arithmetic processing device such as a CPU (Central Processing Unit), an MPU (Micro-Processing Unit), and a GPU (Graphics Processing Unit), and performs various information processing, control processing, and the like related to the information processing device 1 by reading and executing a control program 1P (program product) stored in the storage unit 12. The control program 1P can be deployed so as to be executed on a single computer or at one site, or distributed across multiple sites and on multiple computers interconnected by a communication network. The control program 1P may be read from a portable storage medium 1a and stored in the storage unit 12. In FIG. 1, the control unit 11 is described as being a single processor, but it may be a multiprocessor.

[0015] The storage unit 12 includes memory elements such as RAM (Random Access Memory) and ROM (Read Only Memory), and stores a control program 1P, a tile detection model M1, and a tile position discrimination module M2 necessary for the control unit 11 to execute processing. Details of the tile detection model M1 and the tile position discrimination module M2 will be described later. The tile detection model M1 or the tile position discrimination module M2 may be stored in an external server (not shown). In this case, the control unit 11 of the information processing device 1 may read out the tile detection model M1 or the tile position discrimination module M2 from the external server, or may transmit data to be input to the tile detection model M1 or the tile position discrimination module M2 to the external server and acquire the data output by the external server.

[0016] The input unit 13 accepts various instructions input by the user. In this embodiment, the user is one of the players playing mahjong, who has won a hand of tiles. The display unit 14 displays a photography instruction screen that accepts an instruction to photograph the winning tile, a score display screen that displays the score at the time of winning, and the like. In this embodiment, the information processing device 1 is a smartphone, and the input unit 13 and the display unit 14 are integrated into a touch panel.

[0017] The photographing unit 15 photographs a winning image including a winning tile in a mahjong game. The function of the photographing unit 15 according to the present embodiment is realized by a camera built into a smartphone. Note that the function of the photographing unit 15 may be realized by a photographing device connected to the information processing device 1 by wire or wirelessly.

[0018] Figure 2 is an explanatory diagram showing the winning hand. Mahjong is played using four mahjong tiles with 34 different designs: Man tiles (1 Man to 9 Man), Ton tiles (1 Ton to 9 Ton), So tiles (1 So to 9 So), Wind tiles (East, South, West, North), and Three Elements tiles (White, Fa, and Center). For example, one of the four Go Man, Go Ton, and Go So tiles is a red Dora tile. Red Dora tiles are different from the other tiles in that the entire pattern is painted red.

[0019] 2A shows a winning hand (no concealed kong) in front of a player. The winning hand (no concealed kong) in front of a player includes 13 tiles arranged horizontally (left to right) so that the distance between each tile is less than a first predetermined distance (e.g., the same length as the short side of the tile), and a winning tile that is placed vertically (up to down) away from the 13 tiles arranged horizontally by a second predetermined distance (e.g., the same length as the long side of the tile) or more.

[0020] FIG. 2B shows a winning hand in front of a player (with a concealed kong). When the winning hand includes one concealed kong, the winning hand in front of a player includes ten tiles in front of the player, a concealed kong consisting of four tiles placed horizontally at a position equal to or more than a first predetermined distance from the right edge of the tiles in front of the player, and the winning tile. At least some of the concealed kong are placed face down (with the side with the design on it facing down). The number of tiles included in the winning hand in front of a player changes depending on the number of concealed kongs.

[0021] FIG. 2C shows a winning hand with exposed tiles. The winning hand with exposed tiles includes the front hand tiles arranged horizontally so that the distance between each tile is less than a first predetermined distance, an exposed tile consisting of three tiles placed horizontally at a position that is a first distance or more away from the right end of the winning hand, and the winning tile. The number of tiles included in the front hand varies depending on the number of exposed tiles. The exposed tiles are a triplet, a sequence, or a kong. When the exposed tiles are a kong, the exposed tiles consist of four tiles. FIG. 2C shows a case where there are two exposed tiles. When there are multiple exposed tiles, the exposed tiles are placed horizontally. The exposed tiles include one tile placed in a direction in which the long side extends horizontally.

[0022] FIG. 3 is an explanatory diagram for explaining the tile detection model M1 and the tile position discrimination module M2. The tile detection model M1 is a model that has a segmentation function such as RCNN (Regions with Convolutional Neural Network), Fast RCNN, Faster RCNN, SSD (Single Shot Multibook Detector), YOLO (You Only Look Once), or Vision Transformer, and detects objects. When the tile detection model M1 is configured with a neural network including a CNN (Convolutional Neural Network) that extracts image features such as RCNN, the tile detection model M1 has a plurality of neurons that accept input of pixel values ​​of the winning image, and passes the input pixel values ​​to the intermediate layer. The intermediate layer has a plurality of neurons that extract image features of the winning image, and passes the extracted image features to the output layer. The output layer outputs the type, angle, and position of each tile included in the winning image based on the image features. The control unit 11 may output the type and position of each tile by a method such as pattern matching.

[0023] In the example shown in the image of the detection results in Figure 3, the type and accuracy of the output tile are displayed above each tile. In the example shown in Figure 3, m indicates a Man tile, p indicates a Tong tile, and s indicates a Su tile. If the detected tile is a red Dora tile, it is indicated by the addition of r. Tile detection model M1 detects Man tile, Tong tile, Su tile, Wind tile, and three element tiles, as well as tiles that are laid down, known as "hidden tiles."

[0024] The control unit 11 inputs the winning image in which the type of each tile is output to the tile position discrimination module M2. When the winning image in which the type of each tile is output is input, the tile position discrimination module M2 specifies the position of each tile included in the winning image, and specifies the front hand tile, the exposed tile, and the winning tile based on the specified positions.

[0025] Specifically, when a winning image is input, the tile position discrimination module M2 measures the length (e.g., the number of pixels) of the short side and the long side of each tile in the winning image. The tile position discrimination module M2 averages the measured short side and long side lengths of each tile to calculate the average short side length and the average long side length, respectively.

[0026] The tile position discrimination module M2 identifies the tiles positioned horizontally from the bottom leftmost tile with a distance between each tile less than a first predetermined distance (for example, a distance equal to the average short side length) as tiles in front of the player.

[0027] The tile position discrimination module M2 identifies three tiles that are located at a position more than a first predetermined distance laterally (to the right) from the rightmost tile among the tiles included in the front hand and are arranged at intervals less than the first predetermined distance as exposed side tiles (exposed side tiles of the front hand). Note that, when four tiles are arranged at a position more than a first predetermined distance from the front hand tiles and are arranged at intervals less than the first predetermined distance, the tile position discrimination module M2 identifies the four tiles as exposed side tiles of a Kong. Also, when four tiles are arranged at a position more than a first predetermined distance from the front hand tiles and are arranged at intervals less than the first predetermined distance, and at least one of the four tiles is a concealed tile, the tile position discrimination module M2 identifies the four tiles as a concealed Kong (see FIG. 10).

[0028] The tile position discrimination module M2 identifies a tile located at a position vertically (upward) away from the front hand tile by a second predetermined distance (for example, a distance equal to the average long side length) or more as the winning tile.

[0029] 4 is an explanatory diagram showing the shooting instruction screen. The control unit 11 causes the display unit 14 to display the shooting instruction screen. When the control unit 11 causes the display unit 14 to display the shooting instruction screen, it starts up the shooting unit 15. On the shooting instruction screen, a real-time image 21 captured in real time by the shooting unit 15, a own wind dial 22, a playing wind button 23, a ron win button 24, and a tsumo win button 25 are displayed.

[0030] The own wind dial 22 is a dial for inputting the selection of the own wind when winning. The user can input the own wind when winning by rotating the dial and aligning the own wind when winning with the slot indicated by the arrow.

[0031] The prevailing wind button 23 is a dial for inputting the prevailing wind at the time of winning. When the prevailing wind button is pressed, the prevailing wind displayed on the prevailing wind button is switched to "East" or "South". The display of the prevailing wind button 23 when the winning image is taken is input as the prevailing wind at the time of winning.

[0032] When the ron win button 24 or the tsumo win button 25 is pressed, the control unit 11 causes the photographing unit 15 to photograph a winning image. When the user wins with a discarded tile from another player, the user presses the ron win button 24 to cause the information processing device 1 to photograph a winning image, and when the user wins with a tsumo tile, the user presses the tsumo win button 25 to cause the photographing unit 15 to photograph a winning image.

[0033] Fig. 5 is an explanatory diagram showing a score display screen. Fig. 6 is an explanatory diagram showing a Dora input screen. When the photographing unit 15 photographs a winning image, the control unit 11 causes the display unit 14 to display the score display screen shown in Fig. 5. On the score display screen, a winning image 31, an additional role input field 32, a Dora number input button 33, a main role number input field 34, a winning role display field 35, a score display field 36, and a re-photograph button 37 are displayed.

[0034] The winning image 31 on the score display screen displays the type and probability of the tile output by the tile detection model M1, as well as the positions of the front hand tiles, exposed tiles, and winning tiles output by the tile position discrimination module M2.

[0035] The additional role input field 32 displays check boxes (function buttons) for accepting input of the additional roles of Riichi, Double Riichi, Ippatsu, Yarikan, Reijo Kaikai, Kaitei Raotsuki, or Kaitei Raotsuki. The user can input an additional role by pressing the check box of the additional role that corresponds to the winning tile. When an additional role is input, the role displayed in the winning role display field 35, which will be described later, and the winning score displayed in the score display field 36 are changed.

[0036] When the Dora number input button 33 is pressed, the control unit 11 causes the display unit 14 to display a Dora input screen shown in FIG. 6. The Dora input screen includes a Dora selection field 331. The Dora selection field 331 displays the patterns of tiles included in the winning image. The user can input the selection of Dora by pressing the pattern of a Dora tile. When the pattern of a Dora tile is pressed, the number of Dora tiles is input according to the number of tiles in the winning image 31 whose pattern is pressed. In the example shown in FIG. 5 and FIG. 6, the pattern of a "F" tile is pressed in the Dora selection field 331, and three "F" tiles are included in the winning image 31, so the number of Dora tiles is recognized as three. When a completion button for ending the selection of Dora is pressed after the selection of Dora, the control unit 11 causes the display unit 14 to display the score display screen shown in FIG. 5 again. At this time, the number of han of the Dora hand displayed in the winning hand display field 35 (see FIG. 5) described later and the winning points displayed in the point display field 36 are changed. Note that the display unit 14 may display a field for receiving input of the number of Dora cards.

[0037] The number of wins and function buttons to increase or decrease the number of wins are displayed in the number of wins input field 34. The initial value of the number of wins displayed in the number of wins input field 34 is 0. The user can increase or decrease the displayed number of wins by pressing the function buttons to increase or decrease the number of wins (the up arrow function button and the down arrow function button). When the number of wins is increased or decreased, the winning points displayed in the score display field 36 described below are changed.

[0038] The winning hand display field 35 displays the winning hand corresponding to the winning tile and the number of han for each winning hand, which are specified by the control unit 11 based on the type of each tile output by the tile detection model M1. When an additional hand is input in the additional hand input field 32, or when Dora is selected in the Dora selection field 331, the number of han displayed in the winning hand display field 35 is changed.

[0039] The score display field 36 displays the number of points, the number of han, and the winning points determined by the control unit 11 based on the type of each tile output by the tile detection model M1. Note that the number of han and the winning points displayed in the score display field 36 are changed when an additional hand is input in the additional hand input field 32, when a main hand number is input in the main hand number input field 34, or when a Dora is selected in the Dora selection field 331. In addition, when the Ron Win button 24 is pressed on the shooting instruction screen, the score display field 36 displays the total number of points paid by other players who have given up their hand. When the Tsumo Award button 25 is pressed, the score display field 36 displays the total number of points the user will acquire and a breakdown of the points paid by each other player.

[0040] When the re-shoot button 37 is pressed, the control unit 11 causes the display unit 14 to display the shooting instruction screen again, making it possible to shoot the winning image again.

[0041] 7 is a flowchart showing an example of processing of the information processing device 1. The control unit 11 of the information processing device 1 causes the display unit 14 to display a shooting instruction screen (S1). The control unit 11 causes the input unit 13 to accept input of the player's own wind and the field wind (S2). The control unit 11 causes the input unit 13 to accept input of pressing the Ron win button 24 or the Tsumo win button 25 (S3), and causes the photographing unit 15 to photograph a winning image (S4). The control unit 11 inputs the photographed winning image to the tile detection model M1 (S5), and outputs the type and probability of each tile (S6). The control unit 11 inputs the winning image in which the type and angle of each tile have been output to the tile position discrimination module M2 (S7), and identifies the front hand tile, the exposed tile, and the winning tile (S8). The control unit 11 specifies a winning hand based on the type of each tile, the specified front hand tiles, the exposed tiles, and the winning tile (S9). The control unit 11 calculates a winning score based on the type of each tile and the specified winning hand (S10). The control unit 11 causes the display unit 14 to display a score display screen including a winning image 31, an additional hand input field 32, a Dora number input button 33, a winning hand display field 35, a score display field 36, and a retake button 37 (S11). The control unit 11 causes the input unit 13 to accept input of additional hands including Riichi or Ippatsu, the number of Honba, and Dora (S12). The control unit 11 changes the winning score displayed in the score display field 36 according to the input additional hands and Dora (S13). The control unit 11 determines whether the input unit 13 has accepted input of pressing the retake button 37 (S14). If the input unit 13 receives an input of pressing the re-shoot button 37 (S14: YES), the control unit 11 returns the process to S1. If the input unit 13 does not receive an input of pressing the re-shoot button 37 (S14: NO), the control unit 11 ends the process.

[0042] According to the above configuration and processing, the information processing device 1 can calculate the winning points by photographing a winning image including a winning tile and recognizing the type of the tile included in the photographed winning image. The information processing device 1 can recognize the type of tiles that are commonly used, so it can generally calculate the winning points in mahjong games. Note that the tile position discrimination module M2 according to this embodiment identifies the exposed tile based on its position relative to the front hand tiles, but is not limited to this. The tile position discrimination module M2 can identify the orientation of the tile included in the winning image, and for example, identify the exposed tile as a tile in the face of the opponent, including a tile placed with its long side extending horizontally.

[0043] (Embodiment 2) The winning image according to the second embodiment is an image of the entire mahjong table at the time of winning. The information processing device 1 recognizes the table information including the Dora tile at the time of winning, the main hand, the field wind, the player's wind, and the main hand based on the winning image, and calculates the winning points based on the recognized table information.

[0044] 8 is a block diagram showing an example of the configuration of an information processing device 1 according to embodiment 2. The storage unit 12 of the information processing device 1 according to embodiment 2 stores a tabletop information detection model M3. Details of the tabletop information detection model M3 will be described later. The tabletop information detection model M3 may be stored in an external server (not shown). In this case, the control unit 11 of the information processing device 1 may read out the tabletop information detection model M3 from the external server, or may transmit data to be input to the tabletop information detection model M3 to the external server and acquire the data output by the external server.

[0045] A photographing device C is connected to the information processing device 1 according to the second embodiment. The photographing device C is installed in a position where it can photograph the entire top surface (tabletop) of the mahjong table. A control unit of the information processing device 1 acquires the winning image photographed by the photographing device C, and recognizes the Dora tile and tabletop information from the acquired winning image. Note that the information processing device 1 may include a photographing unit 15 as in the first embodiment, and the photographing unit 15 may capture a winning image including the entire tabletop.

[0046] 9 is an explanatory diagram showing a mahjong table T. An information panel TP is provided in the center of the mahjong table T. The information panel TP is provided with a field wind indicator lamp TP1, a player's wind indicator lamp TP2, and a main field indicator lamp TP3. In addition, mountain tiles, river tiles, and Dora indicator tiles are placed on the mahjong table T.

[0047] The prevailing wind indicator lamp TP1 lights up with the characters "East" or "South." The prevailing wind indicator lamp TP1 indicates the prevailing wind at the time of winning by the character that is lit.

[0048] There is one own wind indicator lamp TP2 on each of the four sides, and the own wind indicator lamp TP2 in front of a player whose own wind is "east" will be lit. In other words, when the lamp in front of the user is lit when winning, the own wind at the time of winning will be "east". When the lamp in the user's upper house (left side) is lit when winning, the own wind at the time of winning will be "south". When the lamp in the user's lower house (right side) is lit when winning, the own wind at the time of winning will be "north". When the lamp opposite the user is lit when winning, the own wind at the time of winning will be "west".

[0049] The home game indicator lamp TP3 has a plurality of linear lighting parts. The home game indicator lamp TP3 indicates the home game when a win occurs by the number of lighting parts that are lit.

[0050] Furthermore, in this embodiment, a Dora indicator tile is placed at a location that is, for example, a first predetermined distance or more horizontally (to the right) from the winning tile and a second predetermined distance or more vertically (upward) from the front hand tile. The Dora indicator tiles include two tiles (a face-up Dora indicator tile on the left and a hidden Dora indicator tile on the right). If Riichi is declared before winning, the two indicator tiles are placed with their figures facing up. If Riichi is not declared before winning, the face-up Dora indicator tile on the left is placed with its figure facing up, and the hidden Dora indicator tile on the right is placed with its figure facing down.

[0051] Fig. 10 is an explanatory diagram illustrating the tile detection model M1 and tile position discrimination module M2 according to the second embodiment. When a winning image is input, the tile detection model M1 according to the second embodiment outputs the type of each tile including the Dora tile. In the example shown in Fig. 10, an area including the front hand tile, concealed kong or exposed tiles, the winning tile, and the Dora tile is excerpted and displayed from the winning image. The tile detection model M1 may also output the type of each mountain tile or river tile included in the winning image.

[0052] The tile position discrimination module M2 in embodiment 2 identifies two tiles that are placed horizontally (to the right) away from the winning tile, for example, by a first predetermined distance or more, and vertically (upward) away from the front hand tile, for example, by a second predetermined distance or more, as Dora indicating tiles.

[0053] FIG. 11 is an explanatory diagram for explaining the tabletop information detection model M3. The tabletop information detection model M3 is a model that has a segmentation function such as RCNN, Fast RCNN, Faster RCNN, SSD, YOLO, or Vision Transformer, and detects objects. When the tile detection model M1 is configured with a neural network including a CNN (Convolutional Neural Network) that extracts image features such as RCNN, the tile detection model M1 has a plurality of neurons that accept input of pixel values ​​of the winning image, and passes the input pixel values ​​to the middle layer. The middle layer has a plurality of neurons that extract image features of the winning image, and passes the extracted image features to the output layer. The output layer outputs the positions of the field wind indicator lamp TP1, the player's wind indicator lamp TP2, and the main field indicator lamp TP3 that are lit in the winning image based on the image features. The control unit 11 may output the positions of the field wind indicator lamp TP1, the player's wind indicator lamp TP2, and the main field indicator lamp TP3 that are lit in the winning image by a method such as pattern matching.

[0054] In the example of the detection result shown in Fig. 11, the area including the information panel TP is displayed in an excerpt, and the lit field wind indicator lamp TP1, the lit field wind indicator lamp TP2, the own wind indicator lamp TP2, and the main field indicator lamp TP3 are shown surrounded by a bounding box. The field wind indicator lamp TP1 is associated with the information of the recognized lit characters ("East" or "South"). The main field indicator lamp TP3 is associated with the number of recognized lit parts, i.e., the number of main fields.

[0055] 12 is an explanatory diagram showing a shooting instruction screen according to embodiment 2. The shooting instruction screen according to embodiment 2 displays a real-time image 21a captured in real time by the camera C, a ron-win button 24, and a tsumo-win button 25. The ron-win button 24 and the tsumo-win button 25 are functional buttons similar to those in embodiment 1.

[0056] 13 is an explanatory diagram showing a score display screen according to the embodiment 2. The score display screen according to the embodiment 2 displays a winning image 31a, an additional winning combination input field 32, a winning winning combination display field 35a, a score display field 36a, and a re-photograph button 37. The additional winning combination input field 32 and the re-photograph button 37 are the same as the additional winning combination input field 32 and the re-photograph button 37 according to the embodiment 1.

[0057] The winning image 31a displayed on the score display screen in the second embodiment displays the type and probability of each tile output by the tile detection model M1, the positions of the front hand tiles, exposed tiles, winning tiles, and Dora tiles output by the tile position discrimination module M2, and the table information output by the table information detection model M3. The winning image 31a displayed on the score display screen in the present embodiment displays an area including the front hand tiles, concealed kong or exposed tiles, winning tiles, and Dora tiles, as well as an area including the information panel TP. The winning image 31a may include the entire top surface of the mahjong table T.

[0058] In the winning combination display field 35a according to the second embodiment, the winning combination corresponding to the winning hand tile and the number of han of each winning combination, which are specified by the control unit 11 based on the type of each tile output by the tile detection model M1 and the positions of the wind indicator lamp TP1 and the player's wind indicator lamp TP2 lit in the winning image output by the tabletop information detection model M3, are displayed. The combinations displayed in the winning combination display field 35a include, for example, a wind combination based on the character of the recognized wind indicator lamp TP1, a player's wind combination based on the position of the player's wind indicator lamp TP2 lit at the time of winning, or a Dora combination based on the recognized Dora indicator tile. The control unit 11 may cause the input unit 13 to accept an input for modifying the winning combination displayed on the display unit 14.

[0059] The score display field 36a according to the second embodiment displays the type of each tile output by the tile detection model M1, and the number of points, the number of han, and the winning point determined by the control unit 11 based on the table information output by the table information detection model M3. The score display field 36a also displays the winning number output by the table information detection model M3, and the winning point is added with the points based on the winning number.

[0060] 14 is a flowchart showing an example of processing of the information processing device 1 according to the second embodiment. The control unit 11 of the information processing device 1 according to the second embodiment causes the display unit 14 to display a shooting instruction screen (S21). The control unit 11 causes the input unit 13 to receive an input of pressing the Ron win button 24 or the Tsumo win button 25 (S22), and causes the photographing device C to photograph a winning image (S23). The control unit 11 inputs the photographed winning image to the tile detection model M1 (S24), and outputs the type and probability of each tile (S25). The control unit 11 inputs the winning image in which the type and angle of each tile are output to the tile position discrimination module M2 (S26), and identifies the front hand tile, the exposed tile, the winning tile, and the Dora display tile (S27). The control unit 11 inputs the winning image to the table information detection model M3 (S28), and outputs the positions of the on-field wind indicator lamp TP1, the player's wind indicator lamp TP2, and the main field indicator lamp TP3 (S29). The control unit 11 recognizes the field wind, the player's wind, and the main field displayed on the mahjong table T at the time of winning based on the positions of the on-field wind indicator lamp TP1, the player's wind indicator lamp TP2, and the main field indicator lamp TP3 that are output (S30). The control unit 11 identifies the winning hand based on the type of each tile, the identified front hand tile, the exposed tile, and the winning tile, as well as the recognized field wind and the player's wind at the time of winning (S31). The control unit 11 calculates the winning score based on the type of each tile, the identified Dora tile, the identified winning hand, and the recognized main field at the time of winning (S32). The control unit 11 causes the display unit 14 to display a score display screen including a winning image 31a, an additional role input field 32, a winning role display field 35a, a score display field 36a, and a re-shoot button 37 (S33). The control unit 11 causes the input unit 13 to accept the input of an additional role (S34). The control unit 11 changes the winning score displayed in the score display field 36 according to the input additional role (S35). The control unit 11 determines whether the input unit 13 has accepted the pressing of the re-shoot button 37 (S36). If the input unit 13 has accepted the pressing of the re-shoot button 37 (S36: YES), the control unit 11 returns the process to S21. If the input unit 13 has not accepted the pressing of the re-shoot button 37 (S36: NO), the control unit 11 ends the process.

[0061] According to the above configuration and processing, the information processing device 1 can calculate the winning points based on the table information displayed on the information panel TP of the mahjong table T or the Dora display tiles placed in the shape of the mahjong table T. The tile position discrimination module M2 according to this embodiment may specify the position and orientation of the river tiles. At this time, for example, if the river tiles include a tile placed in such a direction that the long side of the tile extends horizontally, the control unit 11 may specify Riichi as the winning hand. The tile position discrimination module M2 according to this embodiment may specify the positions of the mountain tiles, the river tiles of other players, and the exposed tiles of other players. At this time, the control unit 11 may specify the additional hand based on the mountain tiles, the river tiles of other players, and the exposed tiles of other players.

[0062] (Embodiment 3) The information processing device 1 according to the third embodiment photographs tiles in hand before a win that have not yet reached a winning state, and identifies the number of tiles in hand or waiting tiles based on the tiles included in the photographed image before a win.

[0063] FIG. 15 is an explanatory diagram showing a pre-winning image. The pre-winning image includes winning front hand tiles that have not yet reached a winning state. The winning front hand tiles include a maximum of 13 tiles. The number of tiles included in the winning front hand tiles varies depending on the number of exposed tiles. Each tile in the winning front hand is not placed with its design facing up (downward tile state), but with its surface on which the design is drawn perpendicular to the top surface of the mahjong table T and facing the outside of the mahjong table T (non-downward tile state). The pre-winning image is an image captured so as to capture the surface of a winning front hand tile that is not down and has a design drawn on it.

[0064] 16 is an explanatory diagram showing a shooting instruction screen according to embodiment 3. The shooting instruction screen according to embodiment 3 displays a real-time image 21 captured in real time by the shooting unit 15 and a button 26 for specifying the number of tiles in hand and a waiting tile. When the button 26 for specifying the number of tiles in hand and a waiting tile is pressed, the control unit 11 causes the shooting unit 15 to shoot a winning image.

[0065] FIG. 17 is an explanatory diagram showing the number of available tiles display screen. The number of available tiles display screen displays a pre-win image 51 and a number of available tiles display field 52. When the photographing unit 15 photographs a pre-win image, the control unit 11 inputs the pre-win image to the tile detection model M1 and outputs the type and probability of each tile in the pre-win hand included in the pre-win image. The pre-win image 51 displayed on the number of available tiles display screen displays the type and probability of each output tile. The control unit 11 specifies the number of available tiles in the pre-win hand based on the type of each output tile. If the specified number of available tiles is equal to or greater than one available tile, the control unit 11 causes the display unit 14 to display the number of available tiles display screen. The specified number of available tiles is displayed in the number of available tiles display field 52.

[0066] 18 is an explanatory diagram showing the waiting tile display screen. On the waiting tile display screen, a pre-win image 51a and a waiting tile display column 53 are displayed. When the number of tiles in hand determined by the control unit 11 based on the type of each output tile is a waiting tile, the control unit 11 determines a waiting tile based on the type of each output tile. A waiting tile is a type of tile that can become a winning tile for a winning tile in front of a winning hand. In the waiting tile display column 53, the fact that the number of tiles in the hand before the win included in the pre-win image 51a is in tenpai and the pattern of the identified waiting tile are displayed.

[0067] FIG. 19 is a flowchart showing an example of the processing of the information processing device 1 according to the third embodiment. The control unit 11 of the information processing device 1 causes the display unit 14 to display a shooting instruction screen (S41). The control unit 11 causes the input unit 13 to accept an input of pressing the number of waiting tiles / waiting tiles specification button 26 (S42), and causes the photographing unit 15 to photograph a pre-win image including the winning front hand tile (S43). The control unit 11 inputs the photographed pre-win image to the tile detection model M1 (S44), and outputs the type and accuracy of each tile (S45). The control unit 11 specifies the number of waiting tiles of the winning front hand tiles based on the type of each output tile (S46). The control unit 11 determines whether the number of waiting tiles of the winning front hand tiles is equal to or greater than one waiting tile (S47). If the winning front tile's number of waiting tiles is equal to or greater than Ikkoten (S47: YES), the control unit 11 causes the display unit 14 to display a waiting tiles display screen (S48) and ends the process. If the winning front tile's number of waiting tiles is not equal to or greater than Ikkoten (S47: NO), i.e., the winning front tile is in waiting status, the control unit 11 identifies a waiting tile based on the type of each output tile (S49), causes the display unit 14 to display a waiting tile display screen showing the identified waiting tile (S50), and ends the process.

[0068] According to the above configuration and processing, the user can know the number of tiles in hand or waiting tiles for the winning hand by photographing the winning tile with the information processing device 1.

[0069] The embodiments disclosed herein should be considered to be illustrative in all respects and not restrictive. The technical features described in each embodiment can be combined with each other, and the scope of the present invention is intended to include all modifications within the scope of the claims and equivalents to the scope of the claims. In addition, the independent claims and dependent claims described in the claims can be combined with each other in all combinations regardless of the reference format. Furthermore, the claims use a format in which a claim references two or more other claims (multiple claim format), but this is not limited to this. A format in which a multiple claim (multi-multi claim) that references at least one multiple claim may be used. [Explanation of symbols]

[0070] 1. Information processing device 1a Removable storage media 1P control program 11 Control section 12 Storage section M1 tile detection model M2 Tile position discrimination module M3 Tabletop Information Detection Model 13 Input section 14 Display section 15. Photography Department 21 Real-time images 21a Real-time image 22 Automatic wind dial 23 Wind Button 24 Ronagari Button 25. Tsumoagari Button 26 Tile identification button 31 Winning Image 31a Agari Image 32 Additional role input field 33 Number of Dora input button 35 Winning Hand Display Column 35a Winning Hand Display Column 36 Point display field 36a Score display field 37 Retake button 51 Pre-winning image 51a Before the win 52 Number of listeners display column 53 Tile display field C. Imaging Device

Claims

1. A winning image is obtained by photographing a winning hand consisting of a plurality of mahjong tiles in a mahjong game; By inputting the winning image into a deep learning model that has been trained to output the types of mahjong tiles contained in the winning image when the winning image is input, the types of mahjong tiles contained in the winning image are output; Calculate the winning points based on the type of each output mahjong tile A program that causes a computer to carry out processing.

2. Calculate the winning points based on the positions of the mahjong tiles in the winning image. The program according to claim 1.

3. Identifying a drawn tile based on the position of each mahjong tile in the winning image; Calculate the winning score based on the identified tsumomo tiles The program according to claim 2.

4. Identifying a revealed tile based on the position of each mahjong tile in the winning image; Calculate the winning score based on the identified extra tiles The program according to claim 2 or 3.

5. displaying a Ron-win button for receiving an input of an instruction to take a photograph of a winning hand by a Ron win, and a Tsumo-win button for receiving an input of an instruction to take a photograph of a winning hand by a Tsumo win; When the ron win button or the tsumo win button is pressed, the winning image is photographed.

3. The program according to claim 1 or 2.

6. After the shot image is taken, the shot image and a function button for accepting a selection of re-taking the image are displayed; When the function button for accepting the selection of re-shooting is pressed, the Ron-win button and the Tsumo-win button are displayed again. The program according to claim 5.

7. The player's own wind is displayed using a dial system, and the player's own wind can be selected when winning from the dial that displays the player's own wind.

3. The program according to claim 1 or 2.

8. Display the function buttons for accepting Riichi or Ippatsu inputs, and the Dora selection field, Change the winning score displayed on the display according to the pressed function button and the selected Dora.

3. The program according to claim 1 or 2.

9. Recognizing the red Dora tile included in the winning image, Calculate the winning score based on the number of red dragon tiles recognized 3. The program according to claim 1 or 2.

10. From the winning image, a Dora display tile or the winning wind, the player's own wind, or the main wind displayed on the mahjong table is recognized; Calculate the winning score based on the recognized Dora tile, the wind in play, your own wind at the time of winning, or the main hand 3. The program according to claim 1 or 2.

11. Identifying the number of tiles in hand or waiting tiles based on the mahjong tiles included in the image before winning 3. The program according to claim 1 or 2.

12. A winning image is obtained by photographing a winning hand consisting of a plurality of mahjong tiles in a mahjong game; By inputting the winning image into a deep learning model that has been trained to output the types of mahjong tiles contained in the winning image when the winning image is input, the types of mahjong tiles contained in the winning image are output; Calculate the winning points based on the type of each output mahjong tile Information processing methods.

13. A winning image is obtained by photographing a winning hand consisting of a plurality of mahjong tiles in a mahjong game; By inputting the winning image into a deep learning model that has been trained to output the types of mahjong tiles contained in the winning image when the winning image is input, the types of mahjong tiles contained in the winning image are output; Calculate the winning points based on the type of each output mahjong tile Control Unit An information processing device comprising:

Images