Game monitoring device

Abstract

A self-contained game monitoring device captures an image of a game table, identifies an object associated with the game, and identifies a value associated with the object. In the game process, the device can detect the betting violation of the player according to the pre-configured game rules. At the end of each game, the device may determine a result of the game (e.g., win / win / push for each bet), and may determine whether an action of the banker (e.g., payment for each bet) coincides with a determination of the device. If an inconsistent action is detected, the device may inform the demerger / supervisor of a potential error.

Description

Background Art

[0001] Casinos usually use a series of cameras to detect fraud or mistakes caused by distributors (dealers), but this is usually a manual process, requiring casino staff to monitor the video feed of the camera. Automated processes have been proposed. For example, U.S. Patent No. 8,172,672 describes a system in which a series of cameras located around the game table send images of game cards on the game table to a remote server device located in the casino management room. The remote server device automatically determines the game win / loss results of the player and the distributor by performing image recognition on the image of the camera. The remote server device also automatically determines the payment to the player through the wireless integrated circuit tag in the game chip. If the dividend is inconsistent with the expected result, the remote server device notifies the distributor and the casino hotel manager. As another example, U.S. Patent No. 10,032,335 uses a series of cameras located around the game table to capture images of game cards and pieces on the game table. Artificial intelligence is used to analyze images to determine whether fraud has occurred. BRIEF DESCRIPTION OF THE DRAWINGS

[0002] Figure 1 is a schematic diagram of a game monitoring device of an embodiment.

[0003] Figure 2 is a block diagram of a game monitoring device of an embodiment.

[0004] Figure 3 is a schematic diagram of an example game environment of an embodiment.

[0005] Figure 4 Schematic diagram of a game table according to an embodiment.

[0006] Figure 5 is an illustration of a gaming monitoring device of an embodiment monitoring two player positions at a gaming table.

[0007] Figure 6 yes Figure 5 A zoomed-in view of the two player positions in the game.

[0008] Figure 7 FIG. 4 is a diagrammatic representation of a stack of sheets according to an embodiment of the present invention.

[0009] Figure 8 Flowchart of a game judgment algorithm of an embodiment.

[0010] Fig. 9 is a flow chart of a method of an embodiment of detecting a card or piece.

[0011] Fig.10 is a flow chart of a method of an embodiment of object recognition.

[0012] Fig.11 is a flow chart of a method of an embodiment of card recognition.

[0013] Fig.12 is a flow chart of a method of an embodiment of card recognition.

[0014] Fig.13 is a flow chart of a method of camera focusing embodiment.

[0015] Fig.14 is a flow chart of a method of an embodiment of image storage.

[0016] Fig.15 is a block diagram illustration of a configuration process of an embodiment. Summary of the invention

[0017] The following embodiments generally relate to a game monitoring device. In one embodiment, a game monitoring device is provided, comprising: a display device, a memory, at least one camera, and a processor; the memory is configured to store game rules; the at least one camera is configured to capture an image of a game table, and the game is played on the game table. The processor is configured to: analyze the images of game objects and pieces on the game table captured by the at least one camera to determine the outcome of the game; and display a notification of the game outcome on the display device. The game monitoring device is a self-contained, independent device because both the analysis and the display occur in the game monitoring device without the need to communicate with a remote device during the game.

[0018] Other embodiments are also possible, and each embodiment can be used alone or in combination. DETAILED DESCRIPTION

[0019] As mentioned above, casinos usually use a series of cameras to detect fraud or mistakes caused by distributors, but it is usually a manual process that requires casino staff to monitor the video feed of the camera. Although automated processes have been proposed, they are usually not adopted due to cost, complexity and incompatibility with the existing infrastructure of the casino. For example, some systems may require network wiring between a remote server and a local camera or other equipment on a gaming table. For an established casino, this may require removing a portion of the ceiling, floor or wall of the casino, if such a transformation is possible. In addition, such a system usually requires several cameras to be set around each gaming table, and a special floor plan may be required. In addition, the use of special pieces with integrated circuits increases cost and complexity. In addition, some systems are used to detect errors or frauds at a later time, and do not provide real-time monitoring of the redemption and collection of the pieces of the distributor or real-time detection of violations of the assertion (bet) in the game.

[0020] The following embodiments may be used to address these issues. In one embodiment, a self-contained game monitoring device is used. In this embodiment, the game monitoring device is "self-contained" in that the camera, processor, and distributor and / or player notification system are all housed in a single device, so that during game play, the game monitoring device does not communicate with a remote server or other device to perform its game monitoring functions (although, for example, updates, game history, and learning files may be communicated outside of game play via a wired or wireless connection).

[0021] Figure 1 The example game monitoring device 100 of this embodiment is shown.It should be understood that this description is only an example, unless clearly stated in the claims, the details of this example should not be understood as claims.For example, although the embodiment will be described below for card games (such as poker or blackjack), it should be understood that the game monitoring device 100 can be used to monitor other types of games, and the game monitoring device 100 can be configured with a game judgment algorithm applicable to other game rules.In addition, the game monitoring device 100 can be used for non-card games (wherein, the game object is a rotating ball, a game roulette wheel, and a dice, etc., and the non-game object is a card).In addition, although these embodiments describe the game played in the casino, it should be understood that the game monitoring device 100 can be used for any suitable place, for example, casino resorts, card rooms, racetracks, riverboat casinos, racetracks, bingo halls or Native American casinos (native American casinos), family games, charity events, etc.

[0022] like Figure 1 As shown, the game monitoring device 100 of this embodiment includes one or more cameras (here, a first pair and a second pair of high-definition cameras 110, 120), a display / notification area / screen / device 130 and a housing 140. In one embodiment, the game monitoring device 100 takes the form of a device similar to a tablet or a smartphone.

[0023] Figure 2 1 is a block diagram of components of the game monitoring device 100 of this embodiment. Figure 2 As shown, in this embodiment, the game monitoring device 100 includes at least one camera (to simplify the drawing, only the first pair of cameras 110 are shown) and the above-mentioned display / notification area 130. The game monitoring device 100 also includes a processor 200 and a memory (computer-readable medium) 210, and the memory 210 can store data, including but not limited to game rules, image data, video data, game progress data, text data, and computer-readable program code, which can be executed by the processor 200 to implement the functions described herein.

[0024] Processor 200 may be a microprocessor that can execute computer-readable program codes (such as firmware) stored in memory 210 (which can store other data) or other computer-readable media in game monitoring device 100. Processor 200 may also be in the form of a pure hardware configuration, such as using a processing circuit, a logic gate, a switch, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or a programmable logic controller. This configuration is also referred to as a processor herein. In addition, multiple processors may be used, such as when a central processing unit (CPU) is used with a graphics processing unit (GPU). Therefore, the term "processor" may refer to one or more processors. The firmware and / or hardware of processor 200 may be configured to perform various functions described below and shown in the flow chart. Processor 200 may also be configured to control a camera, display / notification area 130 (touch screen) and any other components in device 100. In one embodiment, processor 200 takes the form of a Raspberry Pi. Of course, other implementations are also possible.

[0025] The game monitoring device 100 also has a wireless communication interface 220 (e.g., a cellular and / or WiFi interface) and / or a physical input / output port 230 to accept a wired connection or storage device (e.g., a USB drive). Although it is not necessary to connect to an external device / server via these communication channels 220, 230 during game monitoring (because the device 100 is a standalone device), these communication channels 220, 230 can be used to provide game rules, configuration information, training information, software updates, etc. to the device 100, and also serve as a way for the device 100 to output stored historical data. The device 100 may include other components.

[0026] The game monitoring device 100 of the present embodiment is a self-contained, independent device that can analyze the game table image captured by the camera of the device and display notifications about the game results without communicating with a remote device during the game (as mentioned above, it can communicate with the remote device outside the game, for example, to receive various updates). Self-containment avoids the above-mentioned problems with the network system that relies on multiple cameras located around the game table to communicate with the remote server device. However, self-containment also brings some challenges. For example, because all cameras are located in the device 100 itself, it may be important to position it in the device 100 so that it sees the entire relevant field of view of the table 20. To assist in achieving this, in one embodiment, the processor 200 is configured to provide automatic focusing of the camera (for example, using artificial intelligence). In addition, in order to ensure that there are a sufficient number of good images as the basis for analysis, the processor 200 can enable the camera to capture images at a selected high frequency (for example, throughout the game process). In this way, if a card or piece cannot be clearly identified in one image, it can be identified in one of the many other images captured. However, because the memory 210 of the device 100 may be limited, and because all captured images are stored internally to the device 100 because the device 100 is self-contained, it is possible for the memory 210 to run out of space. To address this issue, the processor 200 may use a smart storage algorithm in which the retention time of any given image is based on its use. In addition, because a casino may use multiple devices 100, one for each of its many gaming tables, a centralized batch training and configuration tool may be used to customize each device 100. This will be discussed in more detail below.

[0027] Turning again to the attached figure, Figure 3 An example game environment 10 is shown. Figure 3 As shown, in this example, the game monitoring device 100 is located on a pole 15 on or near a gaming table 20, around which are a number of players 25 (only some are shown to simplify the drawing) and a distributor 30. Figure 4 As shown in more detail in FIG. 1 , on table 20 are various pieces 35 representing player 25 assertions, player's cards 40, dealer's cards 45, community cards 50, and dealer's piece set 55. Of course, the type and number of cards and pieces may vary depending on the game being played.

[0028] like Figure 5 and Figure 6As shown, the game monitoring device 100 is positioned so that the camera in the game monitoring device 100 can see the cards and pieces on the table 20, while the dealer and / or player can see the display / notification area 130 (described below). In this example, the first pair of cameras 110 focuses on the cards and pieces at the third player position (position 3) on the table 20, while the second pair of cameras 120 focuses on the cards and pieces at the fourth player position (position 4) on the table 20.

[0029] Generally speaking, the processor 200 of the game monitoring device 100 utilizes the card and piece images captured by the camera, in combination with its understanding of the rules of the game, to determine the game result, and monitor the payment at the end of the game, and can even assist payment through the display area 130. More specifically, the display area 130 can be used for one or more purposes. One purpose can be to notify the distributor and / or supervisor in real time about the error of potential distributors when paying for the determination or the violation of the player on a specific determination during the game. For example, when the distributor makes a mistake (for example, the incorrect piece quantity is paid to the player), the device 100 can provide a visual alarm (for example, the dashboard warning light on the display area 130) and / or an audible alarm (by the speaker (not shown) in the device 100) to immediately cause the attention of the distributor and / or the floor supervisor. The device 100 can display a "clear" button to eliminate visual and audible alarms, and the floor supervisor can press a "clear button" to reset the device 100 after the distributor's error is corrected, or clear the alarm when an unexpected problem occurs in the device 100 itself. The display area 130 can also provide outcome suggestions for each game in real time (e.g., once the game is over, the win / loss results of the game and their win or loss values ​​can be displayed), which can be used by both players and distributors. In addition, the display area 130 can be used to display the distributor piece count in real time. The device 100 can be configured (e.g., by a supervisor) to enable only some of these features.

[0030] Using the images of cards and pieces captured by the camera of the game monitoring device 100, the processor 200 can use edge computing, artificial intelligence and / or machine learning to monitor the distributor's errors in game judgment and piece payment in the game, and / or help the distributor avoid such errors, which may be caused by stress, negligence or fatigue. The processor 200 can generate suggestions and notifications to the distributor / regulator in real time and display them on the display area 130 of the device 100. The processor 200 can also provide real-time monitoring of human participants in the game violating the rules of the game, such as violating the minimum / maximum number of pieces allowed for each determination, and checking whether subsequent determinations comply with the rules of the game. The processor 200 can also provide analysis, such as game statistics and distributor performance statistics, which is very valuable to casino management.

[0031] Specifically, for example, in a casino table game, the distributor determines who wins the game (whether the distributor wins or other players), and pays the player or clears the player's assertion. In the process of judging who wins the hand of cards and paying the winner, human errors may occur. When the distributor is a new employee or is unfamiliar with the game, or when the distributor is tired, the error rate will increase. It is also possible that the distributor is not good at calculating payments in complex multi-assertion games. These errors are usually not conducive to the casino, because the errors that are not conducive to the player may be discovered and corrected by the player. In addition, before the player leaves the game, the error should be corrected in time. Therefore, the monitoring process is very valuable in identifying / correcting human errors on the spot and evaluating the performance of different distributors of a specific game, so that management can assign a suitable distributor to the game. On the other hand, the casino may not want to make large-scale changes to adapt to the monitoring process. Therefore, this independent, autonomous monitoring device 100 may be very ideal because it provides the casino with the ability to automatically monitor / correct / verify table games without changing the infrastructure. That is, the game monitoring device 100 of these embodiments is a stand-alone device that requires at most minimal integration with the existing infrastructure of the casino. The device 100 is highly portable, easy to install, easy to configure, and easy to update with new game rules. These features provide casino operators with little initial installation and subsequent upgrade disruption, and make the device 100 independent of other existing systems employed by the casino.

[0032] The game monitoring device 100 may also have additional functions. For example, the processor 200 may also implement a data tracking and analysis module that provides historical data for desktop performance and / or distributor performance analysis if the casino chooses to retain data for more analysis. For example, historical data may be accessed from the device 100 via a local-area-network (LAN) computer or a cloud server according to a license granted by casino management. More specifically, the data tracking and analysis module may store data from historical games of a certain time period in the memory 210, including timestamps, cards of players and distributors, assertions and win / loss amounts, and corresponding settlement images thereof. The number of historical pieces of the distributor piece disk may also be recorded. Casino managers may turn on this function to synchronize / export captured data to their information technology (IT) systems (or provide an application programming interface (API) to the data) for further analysis.

[0033] When a human error is detected, an additional field may be added to the data set to indicate the error. By analyzing the data log, the return on investment and effectiveness of the device 100 can be closely monitored. The management of the casino can evaluate the profitability of the game or additional assertions (side bets), the performance of different distributors in the game, fraud detection, and the error rate of the game. At the request of the casino, the device 100 can be configured to connect to a cloud server for more data tagging (e.g., image annotation) and supervised learning and training of neural networks. This helps to improve machine learning algorithms and can also be used to provide automated software upgrades for the device 100.

[0034] In another embodiment (see Fig.15 ), a mobile application or software application executing on a processor 1520 of a computing device 1500 (e.g., a computer, tablet, smartphone, etc.) that is separate from (one or more) devices 100-100" (but may be part of the same local area network) can have a centralized batch training and configuration module 1540, 1550 that uses a camera 1510 (integrated or separate from the computing device 1500) to capture images of casino game objects and pieces as training exercises to configure parameters / settings of one or more devices 100-100". In this way, all devices 100-100" in a casino can be batch initialized according to casino-specific game rules. For example, a casino may need multiple devices 100-100", and different casinos may also have some variations in their game rules. Each casino can use online registration to access the software application and / or mobile application. After verification, the casino can log into their account and define their own configuration / settings for each game. The centralized batch training module 1540 can require the system administrator to present all casino pieces (including special designs) and the face of a deck of playing cards to be captured by the camera 1510. Each piece may require multiple snapshots of different angles and distances on the gaming table. A similar process can be performed on playing cards. Once the training is completed, the computing device 1500 can generate a training results file, which can be broadcast to all independent devices 100-100" or copied to a specific device via the data port 230 (e.g., using a USB drive). In addition, each device 100-100" can have its own user interface (UI) to adjust the settings (e.g., maximum and minimum assertions) for a specific card table. If necessary, the casino supervisor can make local changes.

[0035] The following paragraphs provide example embodiments of the game monitoring device 100 of these embodiments. It should be understood that these are merely examples, and that details given herein should not be construed as claims unless explicitly stated in the claims.

[0036] In one example implementation, the cameras 110, 120 of the game monitoring device 100 take snapshots covering the entire game table 20, and the processor 200 provides artificial intelligence driven automatic zooming to automatically zoom in and out to obtain a high-resolution image of the area of ​​interest of the game table 20. In one embodiment, the processor 200 is configured with a computer vision software module with embedded image analysis. The computer vision software module can be used to perform automatic image recognition on the captured continuous image representation to identify all relevant game objects (e.g., cards, pieces and dice) on the game table 20 and the values ​​of objects of interest (e.g., the suit and level of the cards) using, for example, machine learning algorithms and / or deterministic algorithms. When it is necessary to check the value of the cards, pieces or dice, the computer vision software module can be called through multiple steps in the game judgment algorithm of the processor.

[0037] More specifically, in one embodiment, the computer vision software module of the processor performs automatic image recognition by executing a two-step process. First, the processor 200 uses one or both of the following methods to identify objects of interest on the gaming table 20. In one method, the processor 200 uses an edge detector to detect specific geometric shapes (e.g., the shape of the cards, the shape of the stack of pieces, etc.) in the captured image to identify objects of interest and associate these objects with seating locations, distributors, or communities. The processor 200 can analyze the images captured by the device's camera to identify each player's individual card determination area, the distributor's card area, and the public card area. In different games, the number and location of the player's determination points may be different. An image of the desktop layout can be stored with the correct areas marked.

[0038] In the second method, the processor 200 identifies an object of interest (e.g., a card, piece, dice, etc.) for each identification area. For example, the object of interest may be the cards of the distributor / common cards and the cards in the card area of ​​the player, and the pieces in the determination area. In order for the game judgment algorithm of the processor to work correctly, the object may be associated with the player or the distributor.

[0039] The processor 200 then causes the camera to take a picture of the empty gaming table 20 at the initial calibration time and uses that image as the baseline image. Embedded AI-based image processing can automatically identify the call zone for each designated seat in the baseline image. When a new image of the tabletop is taken, the embedded software in the processor 200 can rotate and rescale the new image so that the table edges match the table edges in the baseline image. Assuming the gaming table is static during gaming, the processor 200 can use the differences between the newly captured image of the tabletop 20 and the baseline image to identify objects of interest. Combined with the boundaries of the call zone for each seating location, the embedded software in the processor 200 can associate the identified objects with individual players. Figure 6 As shown, cards and pieces are identified and associated with two active players. Processor 200 can also use the difference between two consecutive images to identify real-time changes in objects on the table. The above processes can be combined to verify each other's outputs and further improve the accuracy of object recognition.

[0040] Next, the software model in the processor 200 can use image recognition to identify the value of the object of interest. The value of the game card can include the value and suit, the value of the casino piece can be the face value of the piece expressed in the local currency nominal value, and the value of the dice can be the number of dice shown. The processor 200 can use computer vision image recognition to identify the cards and pieces. In card recognition, the processor 200 identifies both the suit and the level. The recognition technology can be based on, for example, the matching technology described in "Playing Card Recognition Using Rotational Invariant Template Matching" published in December 2007 by Zheng and Green from the University of Canterbury in Christchurch, New Zealand. The embedded software can provide a variety of templates related to suits and levels, but if the cards used are very different from conventional playing cards, the processor 200 can choose to perform card learning for each playing card during the initial activation process to generate a new template group dedicated to the casino. The corners of the card can be extracted to perform level matching and suit matching. Flower color can also be used to verify suit identification, especially between spades and hearts.

[0041] For piece identification, the processor can capture a side view of the piece stack, such as Figure 7As shown, and using, for example, a Circle Hough Transform to identify the pieces in the piece pile. Once the piece pile (having one or more pieces) is identified, the processor 200 can detect the edges of the pieces to separate the individual pieces. Next, the processor 200 can use the color and stripe pattern of each piece to identify the piece value. The processor 200 can first perform an initial learning step for all casino pieces. In this article, the quality of the image on the piece pile is very important for the overall accuracy of piece counting and piece value identification. To this end, the processor 200 can cause the camera of the device 100 to zoom in on the piece pile to capture a high-quality picture.

[0042] Each time the computer vision module is called, the processor 200 can obtain assertion information for each player; each player's cards (if any); the dealer's cards (if any); and the community cards (if any). Before the game monitoring device 100 is used for live casino games, supervised learning can be performed on casino cards and all kinds of pieces during initial system setup.

[0043] To improve the accuracy of image recognition, high-quality images can be used. The processor 200 can first detect the area of ​​interest, then use the context of the desktop game 20 to determine which object in the image should remain in focus and automatically adjust the camera angle and zoom settings to focus on the specific area. The processor 200 can then examine the target object to decide whether it is sharp enough. In a short time, the processor's artificial intelligence autofocus system can learn and adjust the camera to keep the target object in full focus.

[0044] To overcome the challenge that partially blocked pieces may affect the accuracy of piece value detection, the processor 200 can use a continuous representation of the image to monitor the entire piece determination process, from the initial piece collection to the final placement of the piece in each determination area. Regardless of whether the piece is blocked, this process can accurately detect the main determination, secondary determination and prompts. The processor 200 can also cache previous determinations to perform intelligent analysis when there is no piece change in the same determination area. The continuous representation may require the processor 200 to store some information images, which may take up considerable storage space. The processor 200 can use an intelligent storage space release algorithm to ensure that the system effectively deletes images that are no longer used.

[0045] In addition to performing the functions of the computer vision software module, the processor 200 may also implement a game determination algorithm to detect the start of a game, record the number of pieces for each player at each assertion point, and check whether the assertions meet assertion rules (e.g., each assertion is within a minimum and maximum number of pieces, if designed to be equal, the value of certain assertions for each game is equal, etc.), detect the end of a game, determine the win / loss / push of each assertion for each active player in the game, look up the payout odds and calculate the theoretical payout for each winning assertion for each player, and determine whether the distributor's actions are consistent with the algorithm output for each assertion active at the gaming table. The distributor's actions may include, but are not limited to, taking the player's assertion item if the player loses, taking no action when the player pushes a card, paying the player for each winning assertion if the computer vision software module detects a payout value, and outputting a warning signal if the actions are inconsistent with its theoretical output.

[0046] A separate game determination algorithm may be designed for each specific casino game. For example, for different games, the number of cards distributed may be different, the order of the cards distributed may be different, the time to determine win / loss may be different, the location of the public cards (if any) may be different, and / or the distributor may not even have a card (e.g., in Mississippi Stud or Cajun Stud). In addition, for the same game, the payout odds may be different in different casinos. Therefore, the game determination algorithm may be implemented for a specific game in a specific casino.

[0047] First, in order to check if it is the start of a new game, a picture of the game table is continuously generated by the camera. The computer vision module analyzes the image to see if there are no cards on the table (by looking for the back or front of the card) and if there is at least one assertion in the designated assertion area. If the game starts, the processor 200 records the assertions and checks if any assertions violate the rules of the game (e.g., the minimum / maximum assertions of the player or the equal payout value between two assertions). If there is a violation, the notification system 130 will notify the distributor if there is an assertion violation and the distributor should ask the player to correct the violation.

[0048] Next, the processor 200 checks whether any cards have been distributed, which indicates that the initial assertion has been set. If no cards have been distributed, the processor 200 performs the above actions again to see if the player can still set a new assertion. If cards have been distributed, the processor 200 checks whether all active players of the assertion have received cards. If not, the processor 200 waits and checks again until all active players have received cards. The processor 2002 then updates the assertions of all active players and records the public card exposure status (i.e., the number of exposed cards and their value / suit, and the number of cards that have not yet been exposed). If there are no public cards, the processor 200 skips all steps related to public cards. If there are public cards, the processor 200 checks whether all public cards have been exposed. If not, the processor waits for all public cards to be exposed and then checks the distributor's cards. The processor 200 waits until the distributor's cards are exposed, or determines that there are no distributor cards in the game. The processor 200 then performs a final update to the assertion and checks whether the assertion is in violation. Finally, the processor 200 determines each asserted win / loss / push on the table and calculates the payout for each win. This is called the theoretical result. For each player, the processor 200 checks whether the distributor's actions are consistent with the theoretical result. If there are any inconsistencies, the notification system 130 is activated to notify the distributor / regulator of the incorrect payout. Otherwise, the current round of the game ends and the processor 200 repeats the above steps for the next round of the game.

[0049] Figure 8-Figure 14 Various flow charts are presented that provide example implementations of the various functions described above. It should be understood that these flow charts are merely examples and other implementations may be used.

[0050] Figure 8 800 is a flowchart of a game determination algorithm 805 of an embodiment. Figure 8As shown, processor 200 determines whether it is game start (action 810). If it is game start, processor 200 records assertion and checks assertion violation (action 815). Then, processor 200 determines whether any cards have been distributed (action 820), if cards have been distributed, processor 200 determines whether all players have received cards (action 825). If all players have received cards, processor 200 updates assertion, checks assertion violation, and records public cards (action 830). Next, processor 200 determines whether all public cards have been exposed (action 835). If all public cards have been exposed, processor 200 determines whether all distributors' cards have been exposed (action 840). If all or no distributors' cards are exposed, processor 200 updates assertion and checks violation (action 845). Processor 200 determines and displays each assertion win, lose and push information on the display area 130 of game monitoring device 100 subsequently. Processor 200 then checks for any inconsistencies (action 855). If an inconsistency is found, processor 200 alerts the distributor / supervisor of the error via display area 120 (action 860). After the error is corrected, the distributor / supervisor presses the "clear" button displayed on screen 130, and in response, processor 200 clears the warning (action 865).

[0051] Fig. 9 900 is a flow chart of a method for detecting an embodiment of a card or piece. Fig. 9 As shown, the processor 200 prepares the approximate location of the judgment area and the card area (action 905). Next, the processor 200 then obtains an overview image of the current table from one or more cameras 110, 120 (action 910). The processor 200 then applies edge detection to the judgment area and the card area to find potential objects (action 915). If the processor 200 finds rectangular corners, the processor 200 infers that the image is a card (action 925). However, if the processor 200 finds circular edges, the processor 200 infers that the image is a piece (action 925).

[0052] Fig.10 1000 is a flowchart of a method of an embodiment for object recognition. Fig.10As shown, the processor 200 prepares a reference image of the game table (action 1005). Next, the processor 200 identifies the assertion area and the public cards (if any) (action 1010). The processor 200 receives a picture of the current game table from the camera (action 1015), and rotates and scales the reference image to match the outline of the game table with the current picture (action 1020). Then, the processor 200 matches the assertion area with the reference image and identifies the active assertion (action 1025). Finally, the processor 200 performs differentiation on the current image and the reference image to identify the cards distributed on the table (action 1030), which is the end of the object recognition process (action 1035).

[0053] Fig.11 1100 is a flowchart of a method for card recognition embodiment. Fig.11 As shown, the processor 200 prepares templates for all casino poker ranks (e.g., 2 to Ace playing cards (Ace)) (action 1105). Then, the processor 200 prepares templates for all casino poker suits (action 1110). The processor 200 then uses edge detection to identify each card and isolates the corners of each playing card face (action 1115). Next, the processor 200 applies template matching to identify the rank of each card (action 1120), and then applies template matching to identify the suit of each card (action 1125), which is the end of the card recognition process (action 1130).

[0054] Fig.12 1200 is a flowchart of a method for card identification embodiment. Fig.12 As shown, the processor 200 prepares templates for all casino pieces (action 1205). The processor 200 then detects pieces near the assertion zone (e.g., using a circular Hough transform) (action 1210). Next, the processor 200 uses edge detection on the piece pile to separate the pieces into individual pieces (action 1215). The processor 200 then uses template matching on the edges of the pieces to identify the values ​​of the individual pieces (action 1220), which is the end of the piece identification process (action 1225).

[0055] Fig.13 1300 is a flow chart of an embodiment method for camera focusing. Fig.13As shown, the processor 200 adjusts the camera to obtain an image that matches the reference desktop image well (action 1305) and identifies objects of interest (e.g., cards, assertions, etc.) (action 1310). The processor 200 then determines whether it has previously stored working camera settings (action 1315). If the working camera settings have been previously stored, the processor 200 applies the previous settings to the camera for repositioning and zooming (action 1320). If the working camera settings have not been previously stored, the processor 200 determines which object on the table to focus on and calculates its position relative to the center of the base image (action 1325). The processor 200 then adjusts the camera angle based on the relative position of the object and adjusts the zoom setting to obtain a high-definition picture (action 1330). The processor 300 then determines whether the object is clearly imaged (action 1335). If the object is not clearly imaged, the processor 200 repeats action 1330. If the object is clearly imaged, the processor 200 stores the camera settings for the current position and moves the camera to the next position.

[0056] Fig.14 1400 is a flow chart of a method of an embodiment for defining long-term (LT), medium-term (MT), and short-term (ST) storage lives (act 1405). Fig.14 As shown, if the processor 200 determines that the image generates a notification (action 1410), the processor 200 stores the image in long-term storage 1420. If the processor 200 determines that the image is used for game judgment (action 1420), the processor 200 stores the image in medium-term storage 1425. If neither of these conditions apply, the processor 200 stores the image in short-term storage 1430, and the image is garbage collected by the garbage collector 1435.

[0057] These embodiments may adopt a variety of alternatives. Some examples of these alternatives are listed below. It should be understood that these are only examples and should not be construed as claims unless explicitly stated in the claims. For example, a non-invasive standalone device can be driven by edge computing and computer vision. The device may include one or more artificial intelligence (AI) powered autofocus cameras, embedded processors (e.g., central processing unit (CPU), graphics processing unit (GPU), tensor processing unit (TPU), etc.), memory chips, and other hardware that supports independent operating systems. The device can use a continuous image capture process to capture real-time images of the entire game table or a portion of the game table at a specific frequency without capturing the player's face to avoid privacy issues. The device may have software that combines computer vision technology with cameras and artificial intelligence to achieve image recognition to identify objects on the game table, including but not limited to card values, dice values, and piece values ​​(even if the pieces are stacked). Pre-trained built-in machine learning modules can be used to retrain models offline using real-time images and objects, and software with a configurable game rule engine can be used to determine whether a specific assertion violates the casino's pre-defined rules, and determine the win / loss of the cards of the participating hands of any table game, as well as calculate the payouts of all winning assertions. Before the distributor collects or distributes the pieces for each player, the display / notification area on the device can display the assertion violation and / or win / loss results and the payout of each assertion in real time. The notification system can also detect the distributor's errors in real time and alert the distributor / manager with predicted results. A physical interface can be provided on the device to allow the distributor to verify / reject / consult the notification. The built-in software module can be used to store historical data of card table performance and / or distributor performance analysis for in-depth analysis. Historical data can be accessed in real time or exported in batches from a local area network (LAN) computer to a cloud server. The built-in data analysis module in the device can determine the profitability of each side bet in each game, so that the casino can decide which side assertion to include in a particular game. Multiple devices may be configured in bulk through a mobile application or a software application running on a processor of a standalone computing device, and some device settings may also be adjusted at the device (eg, by a supervisor).

[0058] It should be understood that all embodiments provided in this detailed description are examples only, and other implementations may be used. Therefore, unless explicitly stated in the claims, the components, architectures, or other details presented herein should not be understood as claims. In addition, it should be understood that components shown or described as "coupled" (or "communicating") to each other may be directly coupled (or communicated) to each other, or indirectly coupled (communicated) to each other through one or more components, which may or may not be shown or described herein.

[0059] The foregoing detailed description is intended to be understood as an illustration of selected forms that the invention can take, rather than a definition of the invention. Only the following claims, including all equivalents, are intended to define the scope of the claimed invention. Finally, it should be noted that any aspect of any embodiment described herein may be used alone or in combination with each other.

Claims

1. A game monitoring device, comprising: Display devices; a memory configured to store game rules; at least one camera configured to capture an image of a gaming table on which a game is being played; as well as A processor configured to: analyzing images of game objects and pieces on the gaming table captured by the at least one camera to determine an outcome of a game; as well as displaying a notification regarding a game result on the display device; The game monitoring device is a self-contained, independent device, because the analysis and display all occur in the game monitoring device, and there is no need to communicate with a remote device during the game.

2. The game monitoring device according to claim 1, wherein: The processor is further configured to: using artificial intelligence to analyze images of the gaming table captured by at least one camera to determine areas of interest on the gaming table; and At least one camera is automatically focused on the region of interest.

3. The game monitoring device according to claim 1, wherein: The at least one camera is also configured to capture a plurality of images of the gaming table throughout a gaming session.

4. The game monitoring device according to claim 3, wherein: The number of images of pieces on the gaming table in the plurality of images is high enough that a piece that is at least partially obstructed in one image is not at least partially obstructed in another image.

5. The game monitoring device according to claim 3, wherein: The processor is further configured to execute a smart storage algorithm, wherein the retention time of each image in the plurality of images is based on its usage.

6. The game monitoring device according to claim 1, wherein: Notification of the outcome of the game includes identification of a winner of the game and / or an determined payout.

7. The game monitoring device according to claim 1, wherein: Notification regarding the game outcome includes notification of distributor error.

8. The game monitoring device according to claim 1, wherein: The processor is also configured to display a notification on the display device that the player has violated an assertion rule.

9. The game monitoring device according to claim 1, wherein: The processor is further configured to display a notification of the distributor piece count on the display device.

10. The game monitoring device according to claim 1, wherein: The processor is also configured to store a history of games monitored by the game monitoring device in the memory for later analysis.

11. The game monitoring device according to claim 1, wherein: The processor is also configured to communicate with the remote device outside of gaming to receive software updates.

12. The game monitoring device according to claim 1, wherein: The processor is also configured to communicate with a remote device outside of game play to receive a training file for identifying game objects and / or pieces.

13. The game monitoring device according to claim 1, wherein: The processor is further configured to communicate with a remote device outside of playing a game to receive the game rules.

14. A method for configuring a plurality of game monitoring devices, the method comprising: On your computing device, do the following: Receiving images of game objects and pieces used in the game; performing a training process to associate images of the game objects and pieces with values; Generating a training result file from the training process, wherein the training result file also includes game rules; and The training results file is communicated to a plurality of game monitoring devices, wherein each game monitoring device is a self-contained, independent device, and each game monitoring device determines and displays notifications regarding game results based on the training results file without communicating with a remote device during game play.

15. The method according to claim 14, wherein: The training result file is transmitted to the multiple game monitoring devices in a batch broadcast manner.

16. The method according to claim 14, wherein: The training result file is transmitted serially to the plurality of game monitoring devices.

17. The method of claim 14, further comprising customizing the game rules.

18. A game monitoring device, comprising: Display devices; at least one video camera; means for receiving a training file for recognizing images of game objects and / or pieces of a game; means for automatically focusing said at least one camera on an area of ​​interest on a gaming table where said game is being played using artificial intelligence; means for determining a game outcome based on a training outcome file and an image of an area of ​​interest on said gaming table; as well as Means for displaying a notification regarding the outcome of the game on the display device.

19. The game monitoring device according to claim 18, wherein: The means for receiving the training file comprises a wireless communication interface.

20. The game monitoring device according to claim 18, wherein: The means for receiving said training file comprises an input / output port.

Images