A passive piece-picking device, a Go-playing robot, and its control method

Abstract

This invention relates to the field of Go robot technology, providing a passive stone-retrieving device, a Go robot, and its control method. The device includes an X-axis movement mechanism, a Y-axis movement mechanism, a Z-axis movement mechanism, and a suction cup mechanism. The X-axis movement mechanisms include a Z-axis movement mechanism, and the suction cup mechanism includes a fixing component, an exhaust solenoid valve, a pressure sensor, a multi-layer suction cup, and a T-junction. The Z-axis movement mechanism is connected to the fixing component, and the T-junction is fixed to the fixing component. The T-junction is connected to the multi-layer suction cup, the pressure sensor, and the exhaust solenoid valve. The Z-axis movement mechanism drives the multi-layer suction cup to move along the Z-axis and, in conjunction with the exhaust solenoid valve, creates a negative pressure within the multi-layer suction cup cavity to adsorb the Go stones. During the stone-retrieval process, the controller determines whether the stone retrieval was successful based on the pressure information uploaded by the pressure sensor. This invention reduces noise, improves portability, and reduces costs, while significantly improving the accuracy and efficiency of stone retrieval.

Description

Technical Field

[0001] This invention relates to the field of Go robot technology, and in particular to a passive stone-picking device, a Go robot, and a control method thereof. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Despite significant progress in the development of Go-playing robot technology, a series of challenges remain, including but not limited to dependence on external power sources, insufficient accuracy and stability in stone selection, poor user experience, limited technological adaptability, low efficiency and success rate in stone selection, lighting issues, and cost. Current technologies have failed to adequately address these problems, limiting the practicality and widespread adoption of Go-playing robots.

[0004] (1) Existing piece-taking mechanisms often rely on external air or power sources, which not only increases the complexity of Go robots, but also limits the portability and application scope of the robots, and also causes energy consumption.

[0005] (2) In terms of user interaction, some robots generate a lot of noise during the process of retrieving the seed, which affects the user interaction experience.

[0006] (3) Cost is also one of the key factors limiting the popularization of Go robots.

[0007] (4) As the usage time increases, inaccurate placement or failure to capture pieces occurs from time to time, which directly affects the smoothness of the game and the user experience.

[0008] (5) Existing piece-picking mechanisms are difficult to meet the needs of Go pieces of different types and materials, which limits the versatility of the robot.

[0009] (6) The efficiency and success rate of piece retrieval are affected by the way the pieces are stacked. The existing piece sorting devices have limited functions and cannot meet the needs of online piece retrieval, resulting in significant difficulties in the piece retrieval process.

[0010] (7) Insufficient light inside the chess box directly affects the accuracy of picking up pieces. Summary of the Invention

[0011] To address the shortcomings of existing technologies, this invention provides a passive stone-retrieving device, a Go robot, and its control method. Through the operation of a mechanical linkage mechanism and an exhaust solenoid valve, the cavity contracts, causing internal air to be expelled, creating negative pressure to adsorb the Go stones. This eliminates the need for external energy, reduces noise, improves portability, and lowers costs. Furthermore, by combining the air pressure information uploaded by the air pressure sensor, the success of stone retrieval can be determined, significantly improving the accuracy and efficiency of stone retrieval.

[0012] To achieve the above objectives, the present invention adopts the following technical solution:

[0013] The first aspect of the present invention provides a passive seed-taking device.

[0014] A passive locator includes an X-axis moving mechanism, a Y-axis moving mechanism, a Z-axis moving mechanism, and a suction cup mechanism. The suction cup mechanism includes a fixing component, an exhaust solenoid valve, a pressure sensor, a multi-layer suction cup, and a three-way valve.

[0015] The Z-axis moving mechanism is connected to the fixing component, and the tee is fixed on the fixing component. The first end of the tee is connected to the multi-layer suction cup, the second end of the tee is connected to the air pressure sensor, and the third end of the tee is equipped with an exhaust solenoid valve. The Z-axis moving mechanism is used to drive the multi-layer suction cup to move along the Z-axis direction and cooperate with the exhaust solenoid valve to form a negative pressure adsorption mechanism in the cavity of the multi-layer suction cup.

[0016] The X-axis moving mechanism, Y-axis moving mechanism, Z-axis moving mechanism, exhaust solenoid valve, and air pressure sensor are all connected to the controller. The controller is used to control the actions of the X-axis moving mechanism and the Y-axis moving mechanism to move the multi-layer suction cup directly above the chess piece, and to control the opening and closing of the solenoid valve, the movement and movement distance of the Z-axis moving mechanism. During the piece retrieval process, the controller determines whether the piece retrieval was successful based on the air pressure information uploaded by the air pressure sensor.

[0017] Furthermore, the X-axis moving mechanism includes a housing, a first motor disposed within the housing, and a roller disposed on one side of the housing. The output shaft of the first motor is connected to a transmission wheel, and the roller is connected to the transmission wheel connected to the output shaft of the first motor via a transmission belt. The rotation of the first motor drives the roller to rotate, and the X-axis moving mechanism drives the Y-axis moving mechanism and the Z-axis moving mechanism together to move along the X-axis direction of the chessboard or chessboard support platform.

[0018] Furthermore, the Y-axis moving mechanism includes a housing, a second motor disposed within the housing, and a synchronous belt; a sliding groove is provided on one side of the housing of the Y-axis moving mechanism, and a slider adapted to the sliding groove is provided on the outer side of the housing of the Z-axis moving mechanism. The slider is connected to the synchronous belt, and the output shaft of the second motor is connected to the synchronous belt through a transmission wheel. The rotation of the second motor drives the synchronous belt to rotate, and the synchronous belt drives the slider to move along the sliding groove, thereby driving the Z-axis moving mechanism to move along the Y-axis direction of the chessboard or the chessboard support platform.

[0019] Furthermore, there are two X-axis moving mechanisms, which are detachably connected to both ends of the Y-axis moving mechanism;

[0020] The Z-axis moving mechanism is detachably mounted on the Y-axis moving mechanism.

[0021] Furthermore, the controller is used to control the movement of the X-axis moving mechanism and the Y-axis moving mechanism when placing a piece. After moving the multi-layer suction cup directly above the placement point, it controls the exhaust solenoid valve to open, allowing air to enter the multi-layer suction cup and restore its original shape. The adsorbed piece separates from the multi-layer suction cup, completing the placement.

[0022] Furthermore, the controller is used to control the actions of the X-axis and Y-axis moving mechanisms during piece removal. After moving the multi-layer suction cup directly above the piece removal point, the controller fully opens the exhaust solenoid valve, and the Z-axis moving mechanism moves downward. The air chamber or air bladder of the multi-layer suction cup is compressed as the Z-axis moving mechanism moves downward. When the Z-axis moving mechanism moves a distance exceeding a set value, the controller fully closes the exhaust solenoid valve, and the surface of the multi-layer suction cup adheres tightly to the adsorbed piece. The controller then moves the Z-axis moving mechanism upward to remove the piece. During the upward movement of the Z-axis moving mechanism, the controller determines whether the air pressure information uploaded by the air pressure sensor is below a threshold. If so, the piece removal is successful.

[0023] Furthermore, the controller is connected to an intelligent image recognition system; the intelligent image recognition system is used to identify the pixel coordinates of each grid line on the chessboard, the category of each piece, and the position of each piece.

[0024] Furthermore, the pixel coordinates of each chessboard grid line are obtained through image distortion correction and reprojection, chessboard reconstruction, Hough line detection, and histogram statistics.

[0025] The category and position of each piece are obtained through a trained piece detection model.

[0026] A second aspect of the present invention provides a Go-playing robot.

[0027] A Go-playing robot includes a controller and a passive stone-picking device as described in the first aspect.

[0028] A third aspect of the present invention provides a control method for a Go-playing robot as described in the second aspect, comprising:

[0029] Control the movements of the X-axis and Y-axis moving mechanisms to move the multi-layer suction cup directly above the chess piece;

[0030] The system controls the opening and closing of the exhaust solenoid valve and the movement and distance of the Z-axis moving mechanism to pick up and drop pieces, and determines whether the piece picking is successful based on the air pressure information uploaded by the air pressure sensor during the picking process.

[0031] Compared with the prior art, the beneficial effects of the present invention are:

[0032] 1. The passive piece-retrieving device of the present invention, through the operation of the mechanical linkage mechanism and the exhaust solenoid valve, the cavity contracts, causing the internal air to be discharged, forming a negative pressure, which adsorbs the chess piece. It does not require external energy, reduces noise, improves portability and reduces cost. Moreover, the success of piece retrieval can be determined by combining the air pressure information uploaded by the air pressure sensor, which significantly improves the accuracy and efficiency of piece retrieval.

[0033] 2. The passive seed-retrieving device of the present invention uses a high-precision synchronous belt and a matching transmission wheel to drive the mechanical linkage mechanism, ensuring no slippage during transmission, achieving precise speed control and position positioning, ensuring the synchronous operation of the seed-retrieving device, and improving the stability and reliability of seed retrieval.

[0034] 3. The passive piece-picking device of the present invention employs a machine learning algorithm, which can continuously optimize the recognition accuracy and adapt to pieces of different colors, shapes and sizes. Attached Figure Description

[0035] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments and descriptions of the invention are used to explain the invention and do not constitute an improper limitation of the invention.

[0036] Figure 1 This is a structural diagram of the suction cup mechanism in Embodiment 1 of the present invention;

[0037] Figure 2 This is a front view of a passive sampling device according to Embodiment 1 of the present invention;

[0038] Figure 3 This is a side view of a passive sampling device according to Embodiment 1 of the present invention;

[0039] Figure 4 This is a front view of the Z-axis moving mechanism according to Embodiment 1 of the present invention;

[0040] Figure 5 This is a side view of the Z-axis moving mechanism of Embodiment 1 of the present invention;

[0041] Figure 6 This is a flowchart of chessboard detection in Embodiment 1 of the present invention;

[0042] Figure 7 This is a schematic diagram of the reconstructed chessboard according to Embodiment 1 of the present invention;

[0043] Figure 8 This is a schematic diagram of a chessboard image according to Embodiment 1 of the present invention;

[0044] Figure 9 This is a histogram showing the number of white pixels along the X-axis in Embodiment 1 of the present invention.

[0045] Figure 10 This is a histogram showing the number of white pixels along the Y-axis in Embodiment 1 of the present invention.

[0046] Figure 11 This is a diagram showing the chess piece detection results of Embodiment 1 of the present invention;

[0047] Figure 12 This is a two-dimensional matrix distribution diagram of chess pieces according to Embodiment 1 of the present invention;

[0048] Figure 13 This is a structural diagram of a Go-playing robot according to Embodiment 2 of the present invention;

[0049] Figure 14 This is a flowchart of a Go robot control method according to Embodiment 3 of the present invention. Detailed Implementation

[0050] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0051] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0052] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0053] In this invention, terms such as "upper," "lower," "left," "right," "front," "back," "vertical," "horizontal," "side," and "bottom" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used only to facilitate the description of the structural relationships of the various components or elements of this invention and do not specifically refer to any component or element in this invention. They should not be construed as limiting the invention.

[0054] In this invention, terms such as "fixed connection," "connected," and "linked" should be interpreted broadly, indicating a fixed connection, an integral connection, or a detachable connection; a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can determine the specific meaning of these terms in this invention based on the specific circumstances, and they should not be construed as limitations on the invention.

[0055] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0056] Example 1

[0057] Embodiment 1 of the present invention provides a passive seed-taking device.

[0058] The passive seed retrieval device provided in this embodiment aims to improve the accuracy and stability of seed retrieval, optimize the user interaction experience, enhance technical adaptability, improve seed retrieval efficiency and success rate, solve the light problem, and reduce costs and save energy and protect the environment.

[0059] This embodiment provides a passive seed-retrieving device, including a monocular camera, a controller, a host computer (equipped with an intelligent image recognition system), a mechanical linkage mechanism, and a suction cup mechanism.

[0060] like Figure 2 and Figure 3 As shown, the mechanical linkage mechanism includes an X-axis moving mechanism 6, a Y-axis moving mechanism 7, and a Z-axis moving mechanism 8.

[0061] There are two X-axis moving mechanisms 6, which are fixedly or detachably connected to the two ends of the Y-axis moving mechanism 7, and the Z-axis moving mechanism 8 is set on the Y-axis moving mechanism 7.

[0062] like Figure 2 and Figure 3 As shown, each X-axis moving mechanism 6 includes a cross-shaped housing, a first motor disposed within the cross-shaped housing, and three rollers disposed on one side of the cross-shaped housing. The three rollers are arranged in an inverted isosceles triangle. The output shaft of the first motor is connected to a transmission wheel. The two upper rollers are connected to the transmission wheel connected to the output shaft of the first motor via a transmission belt. The two upper rollers can be placed on the top surface of the chessboard or chessboard support platform, while the lower roller contacts the bottom surface of the chessboard or chessboard support platform. When the first motor rotates, it drives the rollers to rotate, and the X-axis moving mechanism 6 moves along the X-axis direction of the chessboard or chessboard support platform.

[0063] The Y-axis moving mechanism 7 includes a rectangular housing, a second motor housed within the rectangular housing, and a synchronous belt. The Z-axis moving mechanism 8 adopts a 7-shaped housing. The 7-shaped housing is placed on the rectangular housing. A sliding groove is provided on one side of the rectangular housing of the Y-axis moving mechanism 7, and a slider adapted to the sliding groove is provided on the outer side of the 7-shaped housing. The slider is fixedly connected to or detachably connected to the synchronous belt. The output shaft of the second motor is connected to the synchronous belt through a transmission wheel. The rotation of the second motor drives the synchronous belt to rotate, and the synchronous belt drives the slider to move along the sliding groove, thereby driving the entire Z-axis moving mechanism 8 to move along the Y-axis direction of the chessboard or chessboard support platform.

[0064] The synchronous belt uses a 2GT-232mm tooth profile with a 2mm tooth pitch, featuring a precise circular arc tooth shape and minimal backlash to achieve high-precision positioning and smooth torque transmission. Simultaneously, the tooth profile, tooth pitch, tooth width, and pulley diameter of the drive pulley are matched to the synchronous belt specifications, and suitable materials and surface treatments are used to improve wear resistance and corrosion resistance.

[0065] like Figure 1 and Figure 2 As shown, the suction cup mechanism includes a fixing component 1, an exhaust solenoid valve 2, an air pressure sensor 3, a multi-layer suction cup 4, and a three-way valve 5.

[0066] like Figure 1 As shown, the fastener 1 includes a mounting plate and a box with a bottom but no cover fixed to one side of the mounting plate. The bottom and one side of the box are provided with through holes. The diameter of the through holes is equal to the outer diameter of the tee 5. The tee is set inside the box and is fixedly connected to the box.

[0067] The first end of the tee 5 passes through the through hole at the bottom of the housing, the second end of the tee 5 passes through the through hole on one side of the housing, and the third end of the tee 5 faces the top of the housing.

[0068] After the first end of the tee 5 passes through the through hole at the bottom of the housing, it connects to the first end of the multi-layer suction cup 4.

[0069] The second end of the tee 5 passes through a through hole on one side of the housing and is connected to the air pressure sensor 3.

[0070] The third end of the tee 5 is equipped with an exhaust solenoid valve 2.

[0071] The Z-axis moving mechanism is connected to the top of the fixed part 1.

[0072] like Figure 4 and Figure 5 As shown, the Z-axis moving mechanism includes an inverted 7-shaped housing, a fixed plate, a third motor, a half-turntable, and a connecting rod. The two sides of the fixed plate are fixed to the inner side of the 7-shaped housing, and a through hole is provided in the fixed plate. The third motor is fixed to one side of the fixed plate, and its output shaft passes through the through hole in the fixed plate and is fixedly connected to the center of the half-turntable. The edge of the half-turntable is rotatably connected to the first end of the connecting rod, and the second end of the connecting rod is rotatably connected to the top of the mounting plate of the fixing component 1.

[0073] In one embodiment, the mounting plate of the fastener 1 is provided with a slider on the side near the fixing plate, and the fixing plate is provided with a slide rail that matches the slider on the mounting plate, and the slide rail is perpendicular to the horizontal plane.

[0074] The third motor rotates, causing the half-turntable to rotate. The connecting rod rises, and the slider on the mounting plate slides along the slide rail on the fixed plate, causing the suction cup mechanism to follow and rise. The descent is the reverse action, which is simple and quick.

[0075] The multi-layer suction cup 4 is made of materials such as silicone or rubber. The suction cup is made of a soft and durable material to adapt to different chessboard surfaces.

[0076] During use, the Z-axis moving mechanism moves, causing the fixing part 1 to move downward along the Z-axis. After the second end of the multi-layer suction cup 4 contacts the chess piece, it continues to be compressed. After compression, the exhaust solenoid valve 2 is closed, and the Z-axis moving mechanism moves, lifting the fixing part 1 along the Z-axis. Through the elasticity of the multi-layer suction cup 4 itself, a small negative pressure is generated during the recovery process to adsorb the chess piece. After reaching the designated position, the exhaust solenoid valve 2 is opened, connecting to atmospheric pressure, and the chess piece is released from the multi-layer suction cup 4.

[0077] This embodiment provides a passive piece-retrieving device, the core of which is a multi-layer suction cup 4. The suction cup has a deformable cavity structure inside, which can form or release negative pressure by changing the volume of the cavity. Through the operation of the mechanical linkage mechanism and the exhaust solenoid valve, the contraction of the cavity causes the internal air to be discharged, forming negative pressure and adsorbing the Go pieces. When it expands, external air enters the cavity, releasing negative pressure and thus releasing the Go pieces. The automatic contraction of the cavity of the multi-layer suction cup 4 forms a local negative pressure, realizing the adsorption of Go pieces.

[0078] The contraction and expansion of the multi-layer suction cup 4 are controlled by a series of mechanical linkage mechanisms and exhaust solenoid valves. The mechanical linkage mechanism uses the lever principle and the restoring force of the spring to realize the automatic contraction and expansion of the cavity without the need for external energy, which improves the portability and stability of the passive retrieval device.

[0079] In this embodiment, the mechanical linkage mechanism is designed to allow for fine adjustment to control the force applied when adsorbing and releasing the chess pieces.

[0080] As one implementation method, an adaptive piece-picking strategy is designed: based on the information provided by the intelligent image recognition system, an intelligent algorithm is developed to enable the piece-picking device to adapt to different piece placements and adjust the piece-picking force; and based on the position information provided by the intelligent image recognition system, the position of the piece-picking device is adjusted in real time to ensure accurate piece picking.

[0081] Specifically, the first motor, the second motor, the Z-axis movement mechanism, the exhaust solenoid valve 2, and the air pressure sensor 3 are all connected to the controller, which is also connected to an intelligent image recognition system. This embodiment provides a passive piece-picking device that uses an ESP32 as the underlying master controller to control all motor actuators and solenoid valves. It controls motor movement via one serial communication port and communicates with the RV1106 via another serial port to achieve visual detection and processing communication control. It receives the given piece-picking or piece-placing position from the ESP32, then controls the motor to move to the piece-picking position (picking point) to pick up the piece, to the piece-placing position (placing point) to place the piece, and then back to the piece-picking position, repeating this cycle.

[0082] Among them, the intelligent image recognition system can identify the pixel coordinates of each grid line on the chessboard, the category of each piece, and the position of each piece. It can also identify the position of the chess box and the category and position of the pieces in the chess box.

[0083] In this embodiment, the intelligent image recognition system is configured as follows:

[0084] (1) Before playing chess, check the chessboard. The steps are as follows: Figure 6 As shown, the specific steps include: 1) Image distortion correction and reprojection: The intrinsic parameter matrix and distortion parameters of the camera are obtained through monocular camera calibration. Distortion correction is then applied to the images captured by the camera. The resulting images are then reprojected (projection transformation) to obtain a top-down view of the chessboard, and the local chessboard grid is detected; 2) Reconstructing the chessboard grid lines: If chess pieces are included in the top-down view of the chessboard, chess piece detection is performed. The pixel coordinates of the chess pieces are obtained based on the results of the chess piece detection. The detected chess pieces are then replaced with the crosshairs of the chessboard to facilitate the next step of detection, such as... Figure 7 As shown; 3) Hough line detection: After reconstructing the chessboard, the Hough line detection function in OpenCV is used to detect straight lines (including horizontal and vertical lines) in the image to obtain the chessboard image. The detected horizontal and vertical lines are possible chessboard grid lines. The coordinates of the four corner points of the circumscribed rectangle containing the intersection points of these grid lines are calculated to obtain the largest quadrilateral region enclosed by these Hough lines, such as... Figure 8 As shown; 4) Statistical histogram: Scan the chessboard grid lines (including horizontal and vertical lines) in the chessboard image of the previous step, statistically analyze the histogram, determine the chessboard grid lines, obtain the pixel coordinates corresponding to each grid of the chessboard, that is, the pixel coordinates of the chessboard grid lines, and save the detection results of the chessboard grid lines.

[0085] In this embodiment, the steps for determining the chessboard grid lines are as follows:

[0086] ① A histogram is a graphical representation of the pixel value distribution in a chessboard image. In a binary image, pixel values ​​are only 0 (black) and 255 (white). Each bar on the histogram represents the number of white pixels in a certain column (X-axis histogram) or a certain row (Y-axis histogram) of the image.

[0087] like Figure 9 As shown, the X-axis histogram displays the number of white pixels in each column of the chessboard image. In a chessboard image, if a column contains the vertical lines of the chessboard, the number of white pixels in that column will be significantly higher than in other columns, thus appearing as a tall bar on the histogram.

[0088] like Figure 10As shown, the Y-axis histogram displays the number of white pixels in each row of the image. In a chessboard image, if a row contains the horizontal lines of the chessboard, then the number of white pixels in that row will be significantly higher than in other rows, thus appearing as a tall bar on the histogram.

[0089] ② Using the histogram, we can find several columns and rows where the number of white pixels is significantly higher than in other positions. The positions of these columns and rows are the most likely vertical and horizontal lines in the chessboard grid. Determine the chessboard boundary: First, find the highest peaks in the histogram; these peaks represent the chessboard boundary lines. Determine the grid lines: Within the boundary lines, find several other significant peaks; these peaks represent the internal grid lines of the chessboard. Determine the number of squares within the deformation unit enclosed by the boundary lines and internal grid lines. If the criteria are met, proceed to step ③; otherwise, redetermine the chessboard boundary.

[0090] ③ Calculate the coordinates of each grid cell: The pixel coordinates of each checkerboard cell can be calculated using adjacent grid lines.

[0091] ④ Determine if the grid lines are reasonable, such as whether the spacing and number of grid lines are correct. If they are reasonable, exit the thread; otherwise, the camera re-captures images and re-detects the local chessboard grid.

[0092] Example:

[0093] Suppose the X-axis histogram shows significant peaks in columns 50 and 150, meaning the left and right boundaries of the chessboard are likely between these two lines. If column 75 also has a peak, then this could be the first vertical grid line of the chessboard.

[0094] Similarly, the Y-axis histogram shows significant peaks at rows 100 and 300, meaning the upper and lower boundaries of the chessboard are likely between these two lines. If rows 150 and 250 also have peaks, then these could represent the two horizontal grid lines of the chessboard.

[0095] With this information, the boundaries and internal grid lines of the chessboard can be determined, and the pixel coordinates of each chessboard cell can be calculated.

[0096] Note: Ensure the image path is correct; otherwise, an exception will be thrown. The image should be binary, meaning that pixel values ​​only have two values: 0 and 255.

[0097] The above analysis allows for the efficient reading of binarized images, statistical analysis of the distribution of white pixels, and the creation of histograms, providing an intuitive understanding of the chessboard's position and the pixel coordinates of each grid. This method is highly valuable in image processing and computer vision, particularly for image recognition and pattern recognition tasks.

[0098] (2) During the game, piece detection is performed: The RV1106 uses the VI model to acquire image data from the CSI camera, and processes the image data through the VPSS module (cropping, color space conversion, image distortion correction, etc.). Then, the image data is processed by the NPU processor to detect the type and coordinates of the pieces, obtaining the type and pixel coordinates of the pieces, such as... Figure 11 As shown; Output the distribution of chess pieces on the chessboard: Based on the pixel coordinates of the chess pieces and the pixel coordinates of the chessboard grid detected by the chess pieces, the distribution of the chess pieces on the chessboard is obtained, and a two-dimensional matrix distribution map of the chess pieces is output and stored in the chess notation, as shown. Figure 12 As shown.

[0099] The specific steps for detecting the category and coordinates of chess pieces include: cropping the chessboard area from the image data as the detection area, calculating the projection transformation matrix and applying the projection transformation to obtain a top view of the chessboard, and inputting the top view of the chessboard into the chess piece detection model to obtain the category and coordinates of each chess piece.

[0100] It should be noted that if piece detection is performed during chessboard detection, the chessboard area is the boundary of the chessboard top view.

[0101] The training process for the chess piece detection model is as follows:

[0102] ① Collect video or image data of Go games: crop out the chessboard area from the image or video data as the detection area, calculate the projection transformation matrix and apply the projection transformation to obtain a top view of the chessboard;

[0103] ② Data annotation: Label the chess pieces in the chessboard top view with the type of the chess piece (black or white) and the coordinates of the detection box of the chess piece, and save them to the label file. Each image corresponds to one label file;

[0104] ③ Training the chess piece detection model: Using the YOLOv5S algorithm model provided by Luckfox, with the previous images and labels as input, set the training parameters and run the instructions to train the chess piece detection model;

[0105] ④ Model Conversion: The previous step resulted in a best.pt file. This file is then converted to an ONNX model using tools provided by YOLOv5s. Finally, the ONNX model is converted to a RKNN model usable by RV1106 using conversion tools provided by Luckfox, thus transforming the trained piece detection model into a usable piece detection model.

[0106] A Go game record is a way of documenting Go games, detailing the position of every move made by both players and the final result of the game. Game records are not only a record of games but also an important tool for Go learning and research. Here are some basic concepts of Go game records: Board Layout: Go game records typically begin with a 19x19 grid, the standard Go board size. The intersections on the board are the possible placement positions of stones. Stone Representation: In game records, black stones are usually represented by black dots "●", and white stones by white circles "○". In some modern game records, numerical coordinates are also used to represent the positions of stones. Game Record: A game record documents every move of the game, including the order and position of the stones. The record usually begins with black's first move, followed by white's response, alternating until the game ends. Life and Death Markings: In game records, special markings are sometimes used to indicate the life and death status of certain areas, such as using a square "□" to mark living stones and a circle "○" to mark dead stones. Variations: Game records may include variations, which are possible alternative moves used to demonstrate different tactical options and possible game developments. Commentary and Annotations: High-quality game records usually include commentary and annotations from professional players, which help in understanding the strategies and tactics of the game. Game Record Format: Game records can be recorded in different formats, such as text format and SGF (Smart Game Format). SGF is an internationally standard Go recording format that can include various information such as game records, variations, and commentary. Game Record Analysis: By analyzing game records, players can learn different opening strategies, middlegame tactics, and endgame techniques, thereby improving their Go skills. Historical Games: Famous historical games, such as ancient games or classic games played by modern professional players, are often widely studied and circulated as game records. Personal Game Records: Players can also record their own game records for review and self-analysis to identify highlights and mistakes in the game. Go game records are an important part of Go culture. They not only record the history of the games, but also serve as an important medium for the inheritance of Go skills and the improvement of Go techniques.

[0107] The controller controls the first motor, the second motor, the Z-axis moving mechanism, and the exhaust solenoid valve 2 based on the information identified by the intelligent image recognition system.

[0108] Specifically, the host computer also determines the position to capture or place a piece based on the position of the piece and the next move.

[0109] Specifically, the controller controls the first and second motors to move the multi-layer suction cup 4 directly above the piece, based on the position of the piece being picked up or placed. If picking up a piece, the controller controls the X-axis and Y-axis moving mechanisms to move the multi-layer suction cup directly above the picking point. Then, the controller fully opens the exhaust solenoid valve, and the Z-axis moving mechanism moves downwards. The air chamber or air bladder of the multi-layer suction cup is compressed as the Z-axis moving mechanism moves downwards. When the Z-axis moving mechanism moves a distance exceeding a set value, the controller fully closes the exhaust solenoid valve, and the surface of the multi-layer suction cup adheres tightly to the picked-up piece. The controller then moves the Z-axis moving mechanism upwards to pick up the piece. During the upward movement of the Z-axis moving mechanism, the controller checks if the air pressure information uploaded by the air pressure sensor is below a threshold. If it is, the piece is picked up successfully; otherwise, the piece is picked up unsuccessfully. If the piece is picked up unsuccessfully, the controller controls the first and second motors to adjust the position of the multi-layer suction cup 4 and attempts to pick up the piece again. If it is time to place a piece, control the movement of the X-axis and Y-axis moving mechanisms to move the multi-layer suction cup directly above the placement point. Then, control the exhaust solenoid valve to open, and the multi-layer suction cup will return to its original shape as gas enters. The piece that was attracted will separate from the multi-layer suction cup, and the placement will be completed. At this time, the air pressure information uploaded by the air pressure sensor is atmospheric pressure.

[0110] As one implementation method, the set value of the Z-axis moving distance varies depending on the weight, shape, and material of the chess piece.

[0111] As one implementation method, for chess pieces with different attributes, their height h, diameter R, radius r, and weight w are measured, and the set value of the Z-axis moving mechanism's moving distance L = k((r*w) / (h*R)) is calculated, where k is a set coefficient, and the calculated value is stored.

[0112] The setting value of the Z-axis moving distance determines the force of picking up and placing pieces; optimizing it helps to improve the stability of picking up and placing pieces when facing different pieces.

[0113] This embodiment provides a passive piece-retrieving device that integrates a high-resolution camera and advanced image processing algorithms to accurately identify the positions of pieces on the chessboard, providing operational guidance for the mechanical linkage mechanism and significantly improving the accuracy and efficiency of piece retrieval.

[0114] This embodiment provides a passive piece-picking device that uses machine learning algorithms to continuously optimize recognition accuracy and adapt to pieces of different colors, shapes, and sizes.

[0115] The passive sampling device provided in this embodiment adopts a modular design, which facilitates quick replacement and upgrades, and improves flexibility and maintainability.

[0116] This embodiment provides a passive seed-retrieving device that uses a high-precision synchronous belt and matching transmission wheel to drive the mechanical linkage mechanism, ensuring no slippage during transmission, achieving precise speed control and position positioning, guaranteeing the synchronous operation of the seed-retrieving device, and improving the stability and reliability of seed retrieval.

[0117] This embodiment provides a passive sampling device that emphasizes energy conservation and environmental protection. It selects lightweight and high-strength materials to reduce weight and optimizes the mechanical structure and control strategy to reduce energy consumption.

[0118] The passive piece-picking device provided in this embodiment not only improves the performance and user experience of Go-playing robots, but also emphasizes energy conservation, environmental protection, and economic benefits, and has high market application value and promotion prospects.

[0119] This embodiment provides a passive sampling device that requires no external power source, thus improving portability and stability.

[0120] This embodiment provides a passive piece-retrieving device that significantly improves the accuracy and efficiency of piece retrieval through its intelligent image recognition system and adaptive piece-retrieval strategy. Utilizing intelligent algorithms, the device can identify and adapt to different piece placements and automatically adjust the force of piece retrieval and placement. The intelligent image recognition system provides precise positional information of the pieces, guiding the mechanical linkage mechanism to adjust their position and angle in real time, ensuring accurate piece retrieval every time, thereby significantly improving the efficiency and success rate of piece retrieval.

[0121] Example 2

[0122] Embodiment 2 of the present invention provides a Go-playing robot.

[0123] This embodiment provides a Go-playing robot, such as Figure 13 As shown, the chessboard support platform 9, the chess box 10, the camera 11, the intelligent image recognition system, the controller, and the passive piece-retrieving device described in Embodiment 1 are included.

[0124] This embodiment provides a Go-playing robot that uses the ESP-IDF framework and a standardized serial communication interface.

[0125] The chessboard support platform 9 has slides on both sides, and the rollers of the passive piece-taking device described in Example 1 roll along the slides.

[0126] Two chess boxes 10 are set at one end of the chessboard support platform 9, and a camera 11 is set between the two chess boxes 10.

[0127] This embodiment provides a Go-playing robot equipped with a high-resolution camera and advanced image processing algorithms, enabling it to accurately identify the position and posture of Go pieces on the board under different lighting conditions. The application of light source compensation technology ensures high recognition accuracy even inside the poorly lit Go box, directly improving the accuracy of piece selection. Through the integration of these technologies, the Go-playing robot maintains stable performance in various lighting environments.

[0128] The Go-playing robot provided in this embodiment can also be equipped with light source compensation technology to ensure accurate identification of Go pieces under different lighting conditions.

[0129] The Go-playing robot provided in this embodiment adopts a standardized interface design, which enables seamless connection between different modules and improves compatibility.

[0130] Example 3

[0131] Embodiment 3 of the present invention provides a control method for a Go-playing robot as described in Embodiment 2, such as... Figure 14 As shown, it includes:

[0132] Step 1: Start the program and initialize it.

[0133] The initialization steps include multiple threads: (1) calling the server to initialize the chessboard interface; (2) processing serial port data, i.e., opening the serial port and starting a new thread to read serial port data, such as the robotic arm return signal and the user button signal; (3) pushing image data RTSP, i.e., starting a new thread to push the processed image data through RTSP, and playing it on the host side through VLC to observe the data processing results during the game, which is convenient for debugging; (4) detecting the chessboard grid, i.e., starting a new thread to run the chessboard detection steps mentioned in Example 1, and saving the detection results of the chessboard grid lines (the pixel coordinates of the chessboard grid) to the chess score, which will be used in the main thread; if the detection results of the chessboard meet certain conditions, it can be considered that the detection is accurate, and then the detection results are saved and the thread is exited.

[0134] Step 2: Wait for the user to make a move. If a confirmation signal is received via the serial port, it indicates that the user has made a move. If no signal is received, continue waiting for the user to make a move. After the user has made a move, acquire the image and perform piece detection to obtain a two-dimensional matrix distribution map of the pieces. The piece detection steps are the same as those mentioned in Example 1. Compare the current two-dimensional matrix distribution map of the pieces with the previous two-dimensional matrix distribution map of the pieces stored in the chess record to find the position where the user made a move (i.e., the position where the user made the move). Store the acquired image, the detection result of the pieces in the image (two-dimensional matrix distribution map of the pieces), and the position where the user made a move, i.e., save it to the chess record.

[0135] Step 3: Based on the current two-dimensional matrix distribution map of the chess pieces, determine whether there are any captured pieces, process the pieces on the board according to the captured pieces, and after processing, re-detect the pieces, update the two-dimensional matrix distribution map of the pieces, and save it to the game record.

[0136] In this embodiment, the pieces on the chessboard are processed according to the capture situation. This can be done manually by the user or by controlling a passive piece-capturing device.

[0137] Step 4: Send the two-dimensional matrix distribution map of the chess pieces and the position of the user's move to the server, query KataGo to obtain the machine's next move position, and convert the machine's move position into the move command to be executed by the lower-level machine (controller) (including take a piece from the chess box, take the piece position, go to the move position and place the piece, and place the piece).

[0138] Step 5: After the lower-level machine executes the chess-playing command, it sends a robotic arm return signal and waits for the robotic arm to return to its original position. Upon receiving the robotic arm return signal, it acquires the image again and performs chess piece detection, updates the two-dimensional matrix distribution map of the chess pieces, and checks whether the machine's chess-playing position is correct. If correct, it stores the image acquired this time, the detection results of the chess pieces in the image (two-dimensional matrix distribution map of the chess pieces), and the position of the machine's move, i.e., saves it to the chess record.

[0139] When the lower-level machine executes the chess-playing command, the following control is performed:

[0140] Control the movements of the X-axis and Y-axis moving mechanisms to move the multi-layer suction cup directly above the chess piece;

[0141] The system controls the opening and closing of the exhaust solenoid valve, the movement and distance of the Z-axis moving mechanism, and performs piece picking and dropping. During the piece picking process, the system determines whether the piece picking is successful based on the air pressure information uploaded by the air pressure sensor.

[0142] Step 6: Based on the latest two-dimensional matrix distribution map of chess pieces, detect whether the machine captures any pieces after making a move. If so, process the captured positions on the chessboard. After processing, return to Step 2.

[0143] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A passive sampler device, characterized in that: It includes an X-axis moving mechanism, a Y-axis moving mechanism, a Z-axis moving mechanism, and a suction cup mechanism. The suction cup mechanism includes a fixing component, an exhaust solenoid valve, a pressure sensor, a multi-layer suction cup, and a three-way valve. The Z-axis moving mechanism is connected to the fixing component, and the tee is fixed on the fixing component. The first end of the tee is connected to the multi-layer suction cup, the second end of the tee is connected to the air pressure sensor, and the third end of the tee is equipped with an exhaust solenoid valve. The Z-axis moving mechanism is used to drive the multi-layer suction cup to move along the Z-axis direction and cooperate with the exhaust solenoid valve to form a negative pressure adsorption mechanism in the cavity of the multi-layer suction cup. The X-axis moving mechanism, Y-axis moving mechanism, Z-axis moving mechanism, exhaust solenoid valve, and air pressure sensor are all connected to the controller. The controller is used to control the actions of the X-axis moving mechanism and the Y-axis moving mechanism to move the multi-layer suction cup directly above the chess piece, and to control the opening and closing of the exhaust solenoid valve, the movement and movement distance of the Z-axis moving mechanism to pick up and place the chess piece. During the chess piece picking process, the controller determines whether the chess piece picking is successful based on the air pressure information uploaded by the air pressure sensor. The moving distance of the Z-axis moving mechanism can be set to different values ​​depending on the weight, shape, and material of the chess piece. For chess pieces with different attributes, the height h, diameter R, radius r, and weight w are measured, and the set value of the moving distance of the Z-axis moving mechanism is calculated as L=k((r*w) / (h*R)), where k is a setting coefficient. The calculated value is then stored.

2. The passive sampler device as described in claim 1, characterized in that: The X-axis moving mechanism includes a housing, a first motor disposed inside the housing, and a roller disposed on one side of the housing. The output shaft of the first motor is connected to a transmission wheel, and the roller is connected to the transmission wheel connected to the output shaft of the first motor via a transmission belt. The rotation of the first motor drives the roller to rotate, and the X-axis moving mechanism drives the Y-axis moving mechanism and the Z-axis moving mechanism together to move along the X-axis direction of the chessboard or chessboard support platform.

3. The passive sampler device as described in claim 1, characterized in that: The Y-axis moving mechanism includes a housing, a second motor disposed within the housing, and a synchronous belt. A sliding groove is provided on one side of the housing of the Y-axis moving mechanism, and a slider adapted to the sliding groove is provided on the outer side of the housing of the Z-axis moving mechanism. The slider is connected to the synchronous belt, and the output shaft of the second motor is connected to the synchronous belt through a transmission wheel. The rotation of the second motor drives the synchronous belt to rotate, and the synchronous belt drives the slider to move along the sliding groove, thereby driving the Z-axis moving mechanism to move along the Y-axis direction of the chessboard or chessboard support platform.

4. The passive sampler device as described in claim 1, characterized in that: There are two X-axis moving mechanisms, which are detachably connected to both ends of the Y-axis moving mechanism. The Z-axis moving mechanism is detachably mounted on the Y-axis moving mechanism.

5. The passive sampler device as described in claim 1, characterized in that: The controller is used to control the movement of the X-axis and Y-axis moving mechanisms when placing a piece. After moving the multi-layer suction cup directly above the placement point, it controls the exhaust solenoid valve to open, allowing air to enter the multi-layer suction cup and restore its original shape. The piece is then separated from the multi-layer suction cup, completing the placement.

6. The passive sampler device as described in claim 1, characterized in that: The controller is used to control the movement of the X-axis and Y-axis moving mechanisms during piece removal. After moving the multi-layer suction cup directly above the piece removal point, the controller fully opens the exhaust solenoid valve, and the Z-axis moving mechanism moves downward. The air chamber or air bladder of the multi-layer suction cup is compressed as the Z-axis moving mechanism moves downward. When the Z-axis moving mechanism moves a distance exceeding a set value, the controller fully closes the exhaust solenoid valve, and the surface of the multi-layer suction cup adheres tightly to the piece being suctioned. The controller then moves the Z-axis moving mechanism upward to remove the piece. During the upward movement of the Z-axis moving mechanism, the controller determines whether the air pressure information uploaded by the air pressure sensor is below a threshold. If so, the piece removal is successful.

7. The passive sampler device as described in claim 1, characterized in that: The controller is connected to an intelligent image recognition system; the intelligent image recognition system is used to identify the pixel coordinates of each grid line on the chessboard, the category of each piece, and the position of each piece.

8. The passive sampler device as described in claim 7, characterized in that: The pixel coordinates of each chessboard grid line are obtained through image distortion correction and reprojection, chessboard reconstruction, Hough line detection, and histogram statistics. The category and position of each piece are obtained through a trained piece detection model.

9. A Go-playing robot, characterized in that: Includes a controller and a passive sampling device as described in any one of claims 1-8.

10. The control method for a Go-playing robot as described in claim 9, characterized in that: Control the movements of the X-axis and Y-axis moving mechanisms to move the multi-layer suction cup directly above the chess piece; The system controls the opening and closing of the exhaust solenoid valve, the opening degree of the exhaust solenoid valve, and the movement and distance of the Z-axis moving mechanism to pick up and drop pieces. During the picking process, the system determines whether the picking is successful based on the air pressure information uploaded by the air pressure sensor.

Images