KR102519798B1 - Experimental baseball-simulation service system using LiDAR sensor and AI - Google Patents

Abstract

The present invention relates to an experiential baseball-simulation service system using LiDAR sensors and artificial intelligence in which data collected through the LiDAR sensors installed in an actual baseball stadium is used to generate abilities of each player and simultaneously to reflect the generated abilities of each player in a baseball simulation and the baseball-simulation is operated based on the movement of an object detected through the LiDAR sensors installed in an experiential baseball practice field such that the interest, realism, and participation of an experiential baseball game can be increased by increasing the reality and the efficiency and effectiveness of the baseball practice of a player can be maximized by providing comment information on the pitching or hitting of the player.

Description

Experimental baseball-simulation service system using LiDAR sensor and AI}

The present invention relates to an experiential baseball-simulation service system using LiDAR sensors and artificial intelligence, and more specifically, by using data collected through LiDAR sensors installed in actual baseball stadiums, the ability of each player is measured. At the same time as generating, reflecting the generated player's ability in baseball-simulation, and operating baseball-simulation based on the movement of objects detected through lidar sensors installed in the experience-type baseball driving range, it enhances the reality of the experience-type baseball game. An experiential baseball game operating system using LiDAR sensor and artificial intelligence that can increase interest, realism and participation, and maximize the efficiency and effect of baseball practice by providing comment information on the player's pitching or batting. It is about.

LiDAR (Light Detection And Ranging) means measuring the distance by detecting the time difference (Time of flight) or phase difference of the received optical signal after irradiating an optical signal to an external object, and has spatial resolution compared to radar. Due to the advantages of excellent resolution and resolution, it was conventionally used only in special fields such as aviation and satellite fields, but recently, it has been used in various fields such as civil and national fields such as surveillance and reconnaissance, robots, unmanned surface ships, aircraft such as drones, and industrial security and safety fields. is being extended.

LiDAR is widely used for tracking and detecting the trajectory of a detected object because it can model a stereoscopic image in real time by utilizing a three-dimensional point cloud of x, y, and z.

Meanwhile, the popularity of sports simulation experience stadiums where popular sports games such as baseball, soccer, golf, etc. can be enjoyed through simulation indoors or in a specific place is rapidly increasing.

In particular, baseball is one of the most popular sports events not only for actual play but also for viewing, and recently, experience-type baseball simulation devices for general baseball players as well as members of society are receiving great response.

However, the conventional experiential baseball simulation device simply provides a simple situation such as a player hitting a ball launched from a pitching machine or a player throwing a ball toward a strike zone, so the player's interest and participation are very low, and two lines After hitting or pitching by the player using the above infrared sensors, only the direction, angle, and speed of the baseball are detected and displayed, resulting in a very poor reality compared to actual baseball games.

In particular, since baseball is not hitting a ball with a ball placed on the ground like golf, but hitting a baseball that flies to various positions at high speed through a pitching machine, even if a camera sensor capable of auto-focusing or auto-tracking is used, It has a structural limitation in that it cannot detect detailed physical characteristics of a baseball, such as rotation or contact between a baseball and a baseball bat.

In particular, in baseball, not only pitchers have a wide variety of pitches, but even if they are of the same pitch, various variations can be given to the ball depending on the rotation and speed of the ball, and batters batting average depending on the pitch type, speed, direction, etc. It has very different characteristics.

For example, in the case of pitcher 'A' and batter 'A', batter 'B' is very weak against pitcher 'A''s slider pitch, but can be very strong against fastball, and based on the strike zone, pitcher 'B' It is very strong against balls below the inside of A', but can be very weak against balls above the outside.

In other words, in baseball, even if the batter and pitcher are the same, the pitch, speed and

Depending on the direction, very different results can be brought about, but in the conventional experience-type baseball game, these characteristics of baseball, that is, precise fighting and strategy between players are not reflected at all, so the interest and fun of online baseball games are low.

1 is an exemplary view showing a baseball practice device disclosed in Korean Patent Registration No. 10-1744042 (Title of Invention: Sensing device and sensing method used in baseball practice device, and baseball practice device using the same and control method thereof).

The baseball practice device (hereinafter referred to as the prior art) 100 of FIG. 1 is a ball pitching device 110 for pitching a ball B toward a batter's box where a user hits, and an image of the movement of the ball B are continuously photographed and collected, and the collected images are analyzed to extract the ball (B), calculate the coordinates of the ball (B) in the 3D space, and determine the motion model of the ball (B) using the calculated coordinate data. A sensing device 120 that calculates information using the motion model of the ball B determined by doing so, and receives the information calculated by the sensing device 120 for baseball practice or baseball game progress based on the received information. It consists of a control device 130 that implements an image and an image output unit 140 that projects the image processed by the control device 130 onto the screen unit 141 .

In addition, the sensing device 120 is configured to determine a motion model for the ball B that is pitched and moves to the plate by using the initial time value or initial coordinate data of the ball B, and the ball included in the determined motion model ( If the coordinate data of B) is removed from the coordinate data of all balls and the remaining data is more than the preset number, it is determined that the user has hit the pitched ball, and if the remaining data is less than the preset number or does not exist, the user It is configured to judge that it has not hit.

The prior art 100 configured as described above calculates an accurate and fast motion model for what kind of motion the ball B performs, such as calculating the motion parameters of the ball according to pitching or hitting, necessary for the progress of baseball practice / game. It has the advantage of being able to calculate various information accurately and quickly.

In general, camera sensors capable of auto-focusing or auto-tracking not only have a significantly lower detection rate of the ball (B) due to camera focus problems, tracking problems for nearby parts, and problems with being vulnerable to external environments (lighting, negative and negative, etc.). It has a structural limit in which the ball B is sensed only within a specific range (30 to 50 cm).

However, in the prior art 100, as the sensing device 120 is composed of a camera sensor, it does not solve the above-mentioned structural limitations at all, and the accuracy and reliability of calculating the motion model of the ball B are significantly lowered. Since the movement of the ball B corresponding to pitching or hitting is not reflected, the reality of the baseball game is degraded.

In addition, in the prior art 100, since the sensing device 120 is simply installed on the ceiling of the game room and configured to shoot downward, tracking of the trajectory of the ball B on the plane is possible, but the height of the ball B Tracking of the trajectory is impossible, causing a problem in that the accuracy of calculating the motion model of the ball B is further deteriorated.

In addition, in the prior art 100, since the motions of players such as pitcher, batter, catcher, and runner in the baseball simulation image on the screen unit 141 are simply operated according to preset ability values, various situations of actual players in actual baseball games It has a disadvantage in that the reaction and movement are not reflected at all, so the reality and fun are significantly lowered.

In addition, since the prior art 100 does not describe any technique for providing comments on supplementary points, inadequacies, and disadvantages of the pitching and hitting after the pitcher's pitching and the batter's swing, practice efficiency and effectiveness are improved. It has a falling structural limit.

The present invention is to solve this problem, and the problem of the present invention is to generate the ability value for each player by using data collected through LiDAR sensors installed in the actual baseball stadium, and the generated ability value for each player. can be reflected in the baseball-simulation and operating the baseball-simulation based on the movement of the object detected through lidar sensors installed in the experiential baseball driving range, thereby increasing the interest, realism and participation of the experiential baseball game. It is to provide an experiential baseball-simulation service system using lidar sensor and artificial intelligence that can maximize the efficiency and effect of the player's baseball practice by providing comment information on the player's pitching or batting.

The solution of the present invention for solving the above problems is any one of a batting game room that provides a batter-based experiential baseball-simulation to a player and a pitching game room that provides a pitcher-based experiential baseball-simulation to a player. In the experiential baseball-simulation service system including an experiential baseball game operating system that manages the operation of an experiential baseball driving range equipped with in-game rooms: When the experiential baseball game operating system is installed in a game room for batting , A screen on which batter-based baseball-simulation images are output, and when installed in a pitching game room, pitcher-based baseball-simulation images are output; LiDAR sensors that are installed at intervals in vertical and horizontal directions on the top and side of the game room for batting, transmit LiDAR signals to the interior space, and then receive reflected signals; A controller that outputs a batter-based baseball-simulation image to the screen when installed in a batting game room, and outputs a pitcher-based baseball-simulation image to the screen when installed in a pitching game room; It is installed in each at-bat game room and includes a pitching device that throws a baseball toward the at-bat through a throwing ball of a screen of the at-bat game room, and when the screen is installed in the at-bat game room, the throwing ball is on one side When a cube (C2) is a segment obtained by dividing the interior space of the game room into a plurality of pieces, the controller stores location information of each cube (C2); After scanning the lidar signal transmitted and received by the lidar sensors to recognize and track the object of the player and baseball, an object vector information including player motion information and baseball trajectory information is generated. vector information generator; Utilizing a pre-made baseball-simulation program, proceeding with a virtual baseball game based on a virtual stadium that simulates a real baseball stadium, using a preset calculation formula or calculation table to generate the object vector information for practice A baseball-simulation operating unit generating a baseball-simulation image by generating a baseball-simulation image by deriving a result according to object vector information for practice and matching it to a virtual baseball game, and outputting the generated baseball-simulation image to the screen, wherein the controller controls a game at bat When installed in a room, the object vector information generation unit for practice scans the lidar signals transmitted and received by the lidar sensors, and utilizes the scanning information and location information for each cube (C2) to use the lidar image for each cube. a scanning module that generates; With respect to the LIDAR image of each cube C2 generated by the scanning module, a Sobel filter is applied to detect a pixel having a pixel change rate equal to or greater than a threshold value, and an edge image of each cube C2 composed of edge lines is extracted. Edge image conversion module; an object recognition module for recognizing objects including a batter player, a baseball and a bat by analyzing the edge image for each cube (C2) converted by the edge image conversion module for each cube; an object tracking module for tracking motion of the object recognized by the object recognition module; It is executed when the object tracked by the object tracking module is a batter player, and after detecting the movement of each body part of the batter player by utilizing the object tracking information detected by the object tracking module, the detected batter player's a player motion information generation module for generating player motion information including motions of each body part; It is executed when the object tracked by the object tracking module is a bat, and by utilizing the object tracking information detected by the object tracking module, at least one of the recognized bat's moving direction, speed, and rotational trajectory, and the recognized bat's A bat motion information generation module for generating bat motion information by detecting and matching bat collision information (L) including the position and area on the bat where the bat and the baseball collide; It is executed when the object tracked by the object tracking module is a baseball, and by using the object tracking information detected by the object tracking module, at least one of the position, movement speed, trajectory, and firing angle at the time of hitting the recognized baseball is detected. a baseball trajectory information generation module for generating baseball trajectory information including one or more; Object vector information for practice generating object vector information for practice by matching the player motion information, bat motion information, and baseball trajectory information generated by the player motion information generation module, the bat motion information generation module, and the baseball trajectory information generation module It includes a generating module.

delete

delete

In addition, in the present invention, when the controller is installed in the at-bat game room, the baseball-simulation operator interlocks the motion of the pitcher image in the baseball-simulation image with the pitching device so that the baseball B is thrown. It is desirable to control the pitching device.

In the present invention, the baseball-simulation operating unit extracts at least one or more of the speed, rotation, pitch type, and strike zone location for the pitcher's next pitching from the baseball-simulation program, and then matches them to generate pitching-control data to generate the pitching The pitching-control data generated by the device is transmitted, the pitching device includes throwing control means for controlling the rotation, direction and speed of the baseball B during throwing, and the pitching-control from the baseball-simulation operating unit. Upon receiving the data, it is preferable to control the throwing control unit so that the baseball B is thrown according to the pitching-control data received.

In the present invention, the controller further includes a memory for storing positional information of each cube C2 when the segments obtained by dividing the interior space of the corresponding game room into a plurality of pieces are referred to as cubes C2, and the controller When is installed in the pitching game room, the object vector information generation unit for practice scans the LIDAR signals transmitted and received by the LIDAR sensors, and utilizes the scanning information and location information for each cube (C2) set for each cube. A scanning module for generating LiDAR images for each; With respect to the LIDAR image of each cube C2 generated by the scanning module, a Sobel filter is applied to detect a pixel having a pixel change rate equal to or greater than a threshold value, and an edge image of each cube C2 composed of edge lines is extracted. Edge image conversion module; an object recognition module for recognizing objects including a batter player, a baseball and a bat by analyzing the edge image for each cube (C2) converted by the edge image conversion module for each cube; an object tracking module for tracking motion of the object recognized by the object recognition module; It is executed when the object tracked by the object tracking module is a batter player, and after detecting the movement of each body part of the batter player by utilizing the object tracking information detected by the object tracking module, the detected batter player's a player motion information generation module for generating player motion information including motions of each body part; It is executed when the object tracked by the object tracking module is a bat, and by utilizing the object tracking information detected by the object tracking module, at least one of the recognized bat's moving direction, speed, and rotational trajectory, and the recognized bat's A bat motion information generation module for generating bat motion information by detecting and matching bat collision information (L) including the position and area on the bat where the bat and the baseball collide; Executed when the object tracked by the object tracking module is a baseball, and using the object tracking information detected by the object tracking module, the screen-position, which is the position where the recognized baseball collides with the screen, is detected, a baseball trajectory information generating module that generates baseball trajectory information including at least one of the detected screen-position, baseball location, moving speed, and trajectory; It is preferable to include an object vector information generation module for practice that generates object vector information for practice by matching the player motion information generated by the player motion information generation module and the baseball trajectory information generation module with the baseball trajectory information.

In the present invention, when the object recognized from the edge image of each cube (C2) is a player, the object recognition module in the present invention is the center, which is a line connecting median pixels that are intersections of horizontal lines formed equally at regular intervals with the edge line of the player After extracting the line, the player's body is recognized using the extracted center line, and the player movement information generation module uses the movement information of the center line recognized by the object recognition module to obtain weight movement information, which is the central axis of the body. After detecting , it is preferable to generate player movement information so that the detected weight movement information is included.

In addition, in the present invention, the experiential baseball-simulation service system further includes an AI-based operation server and at least one data collection system built in a baseball stadium where an actual baseball game takes place, and the data collection system is a side of the baseball stadium. And second lidar sensors installed at intervals in the horizontal and vertical directions on the upper side, transmitting lidar signals toward the baseball stadium, and then receiving reflected signals; The positional information of each of the cubes C1, which are segments that divide the baseball field into a plurality of pieces, is preset and stored, and by scanning the lidar signal transmitted and received by the second lidar sensors, the radar signal for each cube C1 is stored. After generating this image, a Sobel filter is applied to the LiDAR image for each cube (C1) generated to extract an edge image for each cube (C2), and the extracted edge image for each cube (C1) is based on the AI It includes a controller transmitted to an operation server, and the AI-based operation server analyzes the edge image of each cube (C1) received from the controller to recognize and track an object, and then utilizes the recognized object tracking information to a player-specific ability value information generation unit for generating ability value information for each player; It is executed every second period (T2), and after collecting the ability value information for each player generated by the ability value information generation unit for each player during the second cycle (T2), the collected ability value information for each player is baseball-simulated. It is preferable to include a baseball-simulation management unit that updates the baseball-simulation program to be reflected in the program and transmits the updated baseball-simulation program in conjunction with the baseball-simulation operating unit of the controller.

In addition, in the present invention, the player-specific ability value generation unit includes a data input module that receives edge images for each cube (C1) transmitted from the controller of the data collection system;

an analysis module that analyzes the edge images of each cube C1 inputted by the data input module by utilizing preset location information of each cube C1; After detecting an object including a player, a bat, and a baseball (B) through the analysis data detected by the analysis module, recognizing the type of the detected object and detecting the location information of the object at the same time, recognizing the object. an object recognition/tracking module that tracks the movement of an object; a current state recognition module that is executed when the object recognized by the object recognition/tracking module is a baseball player and recognizes a current state for each player object by utilizing object tracking information input from the object recognition/tracking module; It is executed when the object recognized by the object recognition/tracking module is a batter, and generates batter movement information including the movement of each body part of the batter object, batting type, speed, trajectory and direction of the bat, and batting result. Motion information generating module; It is executed when the object recognized by the object recognition/tracking module is a pitcher, and pitcher movement information including movement of each body part of the pitcher object, pitch type, baseball speed, trajectory and direction, count result information, and batter betting result a pitcher motion information generating module that generates; A fielder motion information generation module that is executed when the object recognized by the object recognition/tracking module is a pitcher and generates fielder motion information including movement of each body part of the fielder object, movement speed, ball catch result, and throwing accuracy. ; a catcher motion information generation module that is executed when the object recognized by the object recognition/tracking module is a catcher and generates catcher motion information including movement of each body part of the catcher object, catch accuracy, and throwing accuracy; Referring to batter motion information, pitcher motion information, fielder motion information, and catcher motion information generated by the batter motion information generation module, the pitcher motion information generation module, the fielder motion information generation module, and the catcher motion information generation module, It includes a capability information generation module for generating capability information for each player, and the current state applied to the current state recognition module is classified into 'attack' and 'defense', and 'defense' categories include 'pitcher' and 'first baseman'. ', 'Catcher', 'Infielder', 'Outfielder' sub-components, and 'Attack' category is sub-component of 'Hitter', 'Full Swing', 'Single hit', 'Cut off', 'Bunt' It is classified into components, and the category of 'infielder' is preferably classified into sub-components of 'preparation before throwing', 'ball catch', and 'throwing'.

In addition, in the present invention, the baseball-simulation operation unit of the controller extracts simulation-result information, which is a result of a player's pitching or batting, from the baseball-simulation program, and at the same time, at least one or more of current ball counting, base running situation, and defensive position. extracts environment information including, and the controller matches the object vector information for practice generated by the object vector information generation unit for practice with the simulation-hitting result and environment information extracted by the baseball-simulation operation unit to request a comment. Comment information request unit for generating data; It is preferable to further include a control unit for transmitting the comment request data generated by the comment information request unit to the AI-based operation server.

In the present invention, the AI-based operating server includes, as input data, object vector information for practice of the comment request data transmitted from the controller, and a database unit storing an optimal motion detection algorithm for outputting optimal pitching or batting information; It is executed every predetermined cycle (T1), and after collecting the comment request data transmitted from the controller during the first cycle (T1), the collected object vector information for practice of each game room, environment information, and simulation-result information are collected. 1) When a player is a batter, detecting the optimal hitting (movement of each body, bat movement including the bat collision position (L)) according to the environment of the batter player in the virtual baseball field of baseball-simulation 2) When a player is a pitcher, optimal pitching (movement of each body, baseball an AI learning unit for learning an optimal motion detection algorithm by deriving a set of parameter values for detecting a ball (B) (trajectory information); It is executed when comment request data is received from the controller, and optimal pitching / hitting information for generating optimal pitching / hitting information by using the received comment request data as input data of the optimal motion detection algorithm learned by the AI learning unit It is preferable to further include an optimum pitching/hitting information generating unit.

In the present invention, the AI-based operation server further includes a comment information generating unit, and the comment information generating unit includes optimal pitching/hitting information generated by the optimal pitching/hitting information generating unit and comment request data transmitted from the controller. a data input module that receives input; The object vector information for practice and the optimal pitching/hitting information input through the data input module are 1) when the player is a pitcher, at least one of 'movement for each body of the player', 'weight movement' and 'type of pitch' 2) a data extraction module for each category that extracts data for each preset category including at least one of 'motion of each body of the player', 'weight movement' and 'bat movement' when the player is a batter; a data comparison module for each category that compares data of the same category of the object vector information for practice and optimal pitching/hitting information extracted by the data extraction module for each category and detects difference data for each category; a visualization module that visualizes the difference data for each category detected by the data comparison module for each category using graphics, pictures, text, or symbols; a comment information generation module for matching difference data for each category visualized by the visualization module, generating comment information including problems and solutions, and then transmitting the generated comment information to the controller; The experiential baseball game operating system preferably further includes a display terminal installed in each game room, and the controller further includes a display unit for displaying comment information received from the AI-based operation server on the display terminal.

According to the present invention having the above problems and solutions, data collected through LiDAR sensors installed in an actual baseball stadium is used to generate a player's ability value and at the same time, the generated player's ability value is reflected in the baseball-simulation. And, by operating a baseball-simulation based on the movement of objects detected through lidar sensors installed in an experiential baseball driving range, it is possible to increase the interest, realism, and participation in an experiential baseball game by increasing reality, and the player's pitching or By providing comment information on batting, the efficiency and effectiveness of the player's baseball practice can be maximized.

1 is an exemplary view showing a baseball practice device disclosed in Korean Patent Registration No. 10-1744042 (Title of Invention: Sensing device and sensing method used in baseball practice device, and baseball practice device using the same and control method thereof).
2 is a configuration diagram showing an experiential baseball-simulation service system using lidar sensor and artificial intelligence, which is an embodiment of the present invention.
FIG. 3 is a conceptual diagram for explaining FIG. 2 .
FIG. 4 is a configuration diagram showing the data acquisition system of FIG. 2 .
Figure 5 is an exemplary view showing the state that the lidar sensors of Figure 4 are installed in a baseball stadium.
FIG. 6 is a plan view for explaining a cube applied to the controller of FIG. 4 .
Figure 7 is a side view of Figure 6;
FIG. 8 is an exemplary diagram for explaining a 3D image for each cube of a trajectory of a ball acquired by the controller of FIG. 4 .
9 is another exemplary view of FIG. 8 .
10 is a block diagram illustrating a controller of FIG. 4 .
11(a) is an exemplary view showing an edge image for each cube extracted by the edge image extraction module of FIG. 4, and (b) is an example view showing a conventional LIDAR image.
FIG. 12 is an exemplary view showing the center line of each cube extracted by the center line extraction module for each cube of FIG. 4 .
13 is a configuration diagram for explaining an experience-type baseball practice field in which the experience-type baseball game operating system of FIG. 2 is built.
FIG. 14 is a block diagram showing the experiential baseball game management system of FIG. 2 .
15 is an exemplary view showing accessories of the batting game room of FIG. 14;
FIG. 16 is an exemplary view showing a game room for pitching in FIG. 14 .
FIG. 17 is an exemplary front view for explaining the cube of the game room used for object detection by the controller of FIG. 14 .
18 is an exemplary side view of FIG. 17;
FIG. 19 is a block diagram illustrating the controller of FIG. 14 .
FIG. 20 is a block diagram showing an embodiment of the object vector information generation unit for practice shown in FIG. 19 .
Figure 21 (a), (b) is an exemplary view for explaining the collision information of the bat and baseball of the present invention.
22 is a block diagram showing a second embodiment of the object vector information generation unit for practice shown in FIG. 19;
23 is a block diagram showing the AI-based operation server of FIG.
24 is an exemplary diagram illustrating a process of detecting motion information of a batter, a pitcher, a fielder, and a catcher from the time when a hit is made in an actual baseball game.
FIG. 25 is a block diagram showing a player-by-player ability value generation unit of FIG. 23 .
FIG. 26 is a block diagram showing the optimal pitching/hitting information generation unit of FIG. 23 .
FIG. 27 is a block diagram illustrating a comment information generation unit of FIG. 23 .

Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

2 is an experiential baseball using an embodiment of the present invention lidar sensor and artificial intelligence - a configuration diagram showing a simulation service system, Figure 3 is a conceptual diagram for explaining FIG.

An experiential baseball-simulation service system (1) using lidar sensor and artificial intelligence, which is an embodiment of the present invention, learns artificial intelligence by using data collected through LiDAR sensors installed in actual baseball stadiums. At the same time, by reflecting the object vector information collected through LiDAR sensors of the experiential baseball driving range in the baseball-simulation, it is possible to increase the interest, realism, and participation in the experiential baseball game by increasing the reality, and to improve the player's pitching or batting. This is to maximize the efficiency and effect of the player's baseball practice by providing comment information.

In addition, the experiential baseball-simulation service system 1 using lidar sensor and artificial intelligence of the present invention, as shown in FIG. Experiential baseball games installed in collection systems 5-1, ..., (5-N) and experiential baseball driving ranges 900 that provide baseball-simulation and experiential baseball game services to players. Operating systems (7-1), ..., (7-N), data collection systems (5-1), ..., (5-N) and experiential baseball game operating system (7-1 ), ..., consists of an AI-based operation server 3 that manages and controls the operation of (7-N) and a communication network 10.

The communication network 10 includes an AI-based operating server 3, a data collection system 5-1, ..., (5-N) and an experiential baseball game operating system 7-1, ..., ( 7-N), and in detail, it can be composed of a wired and wireless network such as a wide area network (WAN), 3G, 4G, 5G, LTE, etc.

Data collection systems 5-1, ..., (5-N) are built in each of the actual baseball stadiums 800.

In addition, the data collection systems 5-1, ..., (5-N) are LiDAR sensors installed at intervals in the baseball stadium 800 in the horizontal and vertical directions to transmit and receive LiDAR signals. and a controller that scans the lidar signals transmitted and received by the lidar sensors to generate lidar images for each preset cube (S1), and then transmits the generated lidar images to the AI-based operation server 3 .

The experiential baseball game operating systems 7-1, ..., and 7-N are built in each of the experiential baseball driving ranges 900 that provide baseball-simulation and experiential baseball game services to players.

At this time, the experience-type baseball driving range 900 includes a plurality of game rooms providing a space where a player pitches or bats.

In addition, the experiential baseball game operating systems (7-1), ..., (7-N) are installed at intervals in the horizontal and vertical directions in each game room, and lidar sensors that transmit and receive lidar signals, and each After operating the baseball-simulation of the game room and scanning the lidar signals transmitted and received by lidar sensors to generate object vector information, the generated object vector information is reflected in the baseball-simulation of the game room, It consists of a controller that transmits the generated object vector information to the AI-based operating server (3).

The AI-based operation server 3 is the data collection system 5-1, ..., (5-N) and experiential baseball game operation system 7-1, ..., (7-N) Manage and control the operation.

Hereinafter, with reference to FIGS. 4 to 25, the data collection system 5-1, ..., (5-N) of the present invention, the experiential baseball game operating system 7-1, ..., (7 -N) and a detailed description of the AI-based operation server 3 will be given.

- Data collection system

4 is a configuration diagram showing the data collection system of FIG. 2, FIG. 5 is an exemplary diagram showing how the lidar sensors of FIG. 4 are installed in a baseball stadium, and FIG. 6 is a diagram for explaining a cube applied to the controller of FIG. It is a plan view for, and FIG. 7 is a side view of FIG.

As shown in FIG. 4, the data collection system 5 is installed at intervals in the horizontal and vertical directions in the baseball stadium 800 where baseball games of actual baseball players are held, and transmits LiDAR signals, Transmitted and received by lidar sensors 53-1, ..., (53-N) and lidar sensors 53-1, ..., (53-N) receiving reflected signals It consists of a controller 51 that analyzes lidar signals and acquires edge images of objects (players, baseballs, and bats) for each preset cube (S1).

LiDAR sensors 53-1, ..., (53-N) are installed apart from each other in the horizontal and vertical directions of the baseball stadium 800 to transmit LiDAR signals, and then transmit reflected signals. be received

At this time, LiDAR (Light Detection And Ranging) fires a laser and measures the distance, concentration, speed, shape, etc. of the object to be measured from the return time, intensity, frequency change, and polarization state change of the scattered or reflected laser. It is similar to RADAR (Radio Detection And Ranging), which measures the properties and calculates the distance by observing the round-trip time to the object using microwaves, but has a difference in that it uses light, unlike radar that uses radio waves. It is also called 'imaging radar'.

In addition, lidar signal information transmitted and received by the lidar sensors 53-1, ..., 53-N is transmitted to the controller 51.

The controller 51 scans the LIDAR signals transmitted and received by the LIDAR sensors 53-1, ..., 53-N and utilizes the point where the light is reflected to use the LIDAR image, which is a 3D modeling image. Acquire

In addition, as shown in FIGS. 6 and 7 , the controller 51 obtains a LiDAR image of an object for each cube S1 by utilizing location information of each of the preset cubes S1 .

At this time, the cubes S1 are formed in the baseball stadium 800 in the horizontal direction, the first horizontal virtual lines L11 spaced apart from each other in the width direction, and formed in the width direction in the horizontal direction, in the longitudinal direction. When divided into second horizontal virtual lines L12 spaced apart from each other and vertical virtual lines L2 formed in the vertical direction and spaced apart from each other in the width and length directions, it means the divided segments.

That is, the controller 51 obtains a lidar image by scanning lidar signals transmitted and received by lidar sensors 53-1, ..., 53-N, and simultaneously acquires the position of the lidar image. LIDAR image for each cube (S1) is obtained by utilizing the information and the location information of each cube (S1), and accordingly, the movement of the baseball player, the baseball (B) and the bat, and the physical It becomes possible to track characteristics precisely and quickly.

FIG. 8 is an exemplary diagram for explaining a 3D image for each cube of a trajectory of a ball obtained from the controller of FIG. 4 , and FIG. 9 is another exemplary diagram of FIG. 8 .

In general, when a pitcher throws the ball, the type, speed and direction of the ball are different depending on various variables such as the position, quantity and shape of the fingers that catch the ball, the strength and direction of the snap of the wrist, and the strength and direction of the swing of the arm. It is classified, and various trajectories of the ball are formed according to the ball type, speed, and direction of the ball.

However, conventionally, since the trajectory of the ball is detected using only the initial position (A) and final position (B) of the baseball (B), the line (L) connecting the initial position and final position of the ball is used to detect the ball's trajectory As a result, the trajectory of the actual ball cannot be accurately reflected, resulting in significantly lowered accuracy and precision.

The present invention can solve this problem, and the controller 51 of the present invention obtains a LIDAR image of an object (a player, a bat or a baseball) for each cube S1, so that the object Since precise tracking is possible, it is possible to precisely track the trajectory of various baseballs according to the type, speed, and direction of the baseball.

For example, as shown in FIG. 8, when a pitcher pitches a curve pitch, conventionally, a line L connecting the initial position (A) and the final position (B) of the ball is simply used as the trajectory of the ball. However, since the controller 51 of the present invention can detect objects by dividing each cube, the trajectory of the ball can be accurately detected in the order of a -> b -> c -> d -> e do.

As another example, as shown in FIG. 9, after detecting the cubes (S11 to S15) including the ball object when a hitter is hit, the positional information of each detected cube and the positional coordinates (x, By using y and z), the trajectory of the ball can be accurately tracked from A -> B -> C -> D -> E.

In other words, the present invention installs a plurality of LiDAR sensors 53-1, ..., (53-N) on the front, side and top of the baseball stadium 800, and at the same time the controller 51 By utilizing the cubes S1 divided by the vertical virtual lines L2 connecting the baseball stadium in the vertical direction and the horizontal virtual lines L1 connecting the horizontal direction, LiDAR images are generated for each cube S1. be able to obtain

In addition, when the lidar image for each cube S1 is generated, the controller 51 analyzes the generated lidar image to extract the edge image of the object (player, bat, and baseball), and then extracts the edge image of the object from the extracted edge image. A center line forming a center line is detected.

FIG. 10 is a block diagram showing the controller of FIG. 4, FIG. 11 (a) is an exemplary view showing an edge image for each cube extracted by the edge image extraction module of FIG. 4, and (b) is a conventional lidar. FIG. 12 is an example view showing a center line for each cube extracted by the center line extraction module for each cube in FIG. 4 .

As shown in FIG. 10, the controller 51 includes a control module 511, a memory 512, a data transmission/reception module 513, a LIDAR signal scanning module 514, and an edge image extraction module 515 for each cube. ), a center line extraction module 516 for each cube, and a data collection module 517.

The control module 511 is an O.S (Operating System) of the controller 51, and manages and controls the control objects 512, 513, 514, 515, 516, and 517.

In addition, the control module 511 transmits the lidar signal information received from the lidar sensors 53-1, ..., (53-N) through the data transmission/reception module 513 to the lidar signal scanning module 514. output as

In the memory 512, installation location information of each lidar sensor 53 and location information for each cube are preset and stored.

In addition, the memory 512 temporarily stores edge images for each cube generated by the edge image extraction module 515 for each cube.

In addition, center line information for each cube generated by the center line extraction module 516 for each cube is temporarily stored in the memory 512 .

The data transmission/reception module 513 transmits and receives data with the lidar sensors 53-1, ..., (53-N) and the AI-based operation server 3.

The lidar signal scanning module 514 scans lidar signals transmitted and received by the lidar sensors 53-1, ..., and 53-N, and at the same time, the position of each of the preset cubes C1. LIDAR images 800 for each cube C1 of FIG. 11(b) are obtained by utilizing the information.

At this time, the lidar image 800 for each cube C1 obtained by the lidar signal scanning module 514 is temporarily stored in the memory 512 under the control of the control module 511, and at the same time, the edge image extraction module for each cube (515) is entered.

The purpose of the edge image extraction module 515 for each cube is to detect the edge of the lidar image 800 for each cube C1 of FIG. 11 (b) obtained by the lidar signal scanning module 514 As shown in FIG. 11(a), an edge image 810 for each cube composed of edge lines is extracted by detecting pixels having a change rate of pixels greater than or equal to a threshold value by applying a commonly used Sobel filter.

At this time, assuming that the controller 51 does not generate the edge image 810 for each cube of FIG. 11 (a), but only the lidar image 800 for each cube of FIG. 11 (b), each cube Since the LiDAR image 800 is composed of a plurality of edges (boundary lines), the amount of data processing is excessively increased, but in the present invention, the edge image extraction module 515 for each cube of the controller 51 is As in a), by generating the edge image 810 for each cube composed of only edges having a rate of change equal to or greater than the threshold value, it is possible to significantly reduce the amount of computational processing and thus maximize computational processing efficiency.

At this time, the edge images 810 for each cube generated by the edge image extraction module 515 for each cube are temporarily stored in the memory 512 under the control of the control module 511, and at the same time, the center line extraction module for each cube ( 516) and data collection module 517.

The center line extraction module 516 for each cube extracts the center lines 101 to 105 of the object from the edge image 810 for each cube extracted by the edge image extraction module 515 for each cube.

In this case, as shown in FIG. 12 , the center lines 101 to 105 refer to lines connecting median pixels of intersections of horizontal lines equally formed at regular intervals with the edge line of the object.

The central line (101 to 105) information for each cube is stored in the AI-based operating server (3), which will be described later, when the detected object is a baseball player, the head (A), hands (B, C), and feet of the baseball player object. It is used to determine the location of (D, E).

In addition, center line information for each cube extracted by the edge image extraction module 516 for each cube is temporarily stored in the memory 512 under the control of the control module 511 and inputted to the data collection module 517 at the same time.

The data collection module 517 extracts the edge image 810 for each cube obtained by the edge image extraction module 515 for each cube and the center line for each cube extracted by the center line extraction module 516 for each cube. After collecting information, location information of each cube (C1) set in advance, and other additional information (baseball stadium location and identification information, controller identification information, etc.), they are matched to generate collected data.

At this time, the collected data generated by the data collection module 517 is transmitted to the AI-based operation server 3 through the data transmission/reception module 513 under the control of the control module 511.

In this way, the data collection systems 5-1, ..., (5-N) of the present invention are built in each of the baseball stadiums 800 where baseball games of actual baseball players take place, and the lidar sensor 53-1 ), ..., (53-N) are configured to detect objects for each preset cube (S1), so that detailed and accurate object tracking and physical characteristics can be obtained from the AI-based operation server (3). .

- Experiential baseball game management system

13 is a configuration diagram for explaining an experience-type baseball practice field in which the experience-type baseball game operating system of FIG. 2 is built.

The experiential baseball game operating system (7-1), ..., (7-N) of the present invention is provided to each of the baseball driving ranges (900) for providing baseball-simulation, experiential baseball game, and teaching services to players. is built

At this time, the experiential baseball driving range 900 is provided with a plurality of game rooms, which are predetermined spaces for providing virtual experiential baseball games and baseball-simulations to players. In detail, as shown in FIG. 13, 1 ) Batting game rooms (910-1), ..., (910-N) for providing players with a batter-based experiential baseball game and baseball simulation, and 2) a pitcher-based experiential baseball game for players and pitching game rooms 920-1, ..., 920-N for providing a baseball simulation.

That is, visitors (hereinafter referred to as players) who visit the experiential baseball practice range 900 may enter the batting game room 910 or the pitching game room 920 according to the type (batter or pitcher) they want to experience the experiential baseball game. By selecting one of them, you can enjoy the type of experiential baseball game you want.

FIG. 14 is a block diagram showing the experiential baseball game management system of FIG. 2 .

As shown in FIG. 14, the experiential baseball game operating system 7 includes accessory equipment installed in the game room 910 for each at-bat and accessory equipment installed in the game room 920 for each pitching. .

At this time, the auxiliary equipment installed in the game room 910 for each at-bat is a controller 71, lidar sensors 72, pitching device 73, display terminal 74, screen 75, video output device ( 76), and the auxiliary equipment installed in each pitching competition room 920 is a controller 71, lidar sensors 72, display terminal 74, screen 75, and video output device 76 ) is made up of

15 is an exemplary view showing accessories of the batting game room of FIG. 14;

The batting game room 910 of FIG. 15 is a game room for providing a batter-based experiential baseball game and baseball-simulation service to players and a teaching service for the player's batting.

In addition, the following accessories are installed in the game room 910 for batting.

As shown in FIGS. 13 and 15, the accessory equipment of the batting game room 910 is installed at intervals in the horizontal and vertical directions on the top and side of the batting game room 910, and LiDAR signals After sending out, the lidar sensors 72 receiving the reflected signal and a screen that is vertically installed in front of the batting game room 910 and displays baseball-simulation images output from the image output device 76 75, an image output device 76 installed on the ceiling of the batting game room 910 to project a baseball-simulation image onto the screen 75, and a player P installed adjacent to the batter's box 911 ) Display terminal 74 for displaying comment information on the batting, a pitching device 73 disposed in the center and front of the screen 75 to throw a baseball B toward the batter's box 911, and tactics It consists of a controller 71 that manages and controls the operation of the constituent means 72, 73, 74, 75, and 76.

The lidar sensors 72 are installed at intervals in the horizontal and vertical directions on the top and side of the batting game room 910.

In addition, the lidar sensors 72 transmit lidar signals to the batting game room 910 and receive reflected signals.

In addition, the lidar sensors 72 transmit the transmitted and received lidar signal information to the controller 71.

That is, since lidar sensors 72 are installed at intervals in horizontal and vertical directions in the batting game room 910 of the present invention, the controller 71 controls the physical characteristics of the player (batter), the bat, and the baseball (B). It can be precisely detected by dividing each cube (C2).

The screen 75 is installed in front of the batter's box 911 and displays baseball-simulation images projected from the image output device 76.

The image output device 76 is a device that projects a baseball-simulated image onto a screen 75 under the control of the controller 71 .

The pitching device 73 is disposed in the center of the screen 75 and is a device that throws the baseball B toward the batter's box 911. At this time, in detail, a throwing ball (not shown) is formed in the center of the screen 75 to pass the baseball B thrown from the pitching device 73, so that the baseball ball B thrown from the pitching device 73 (B) is formed. ) passes through the throwing ball of the screen 75 and moves toward the batter's box 911.

In addition, although not shown in the drawings, the pitching device 73 includes a throwing control means (not shown) for controlling the rotation, direction and speed of the baseball B during throwing, and accordingly, the pitching device 73 When pitching-control data (including strike zone position, speed, rotation, etc.) is received from the controller 71, the player controls the throwing control means so that the baseball (B) is thrown according to the received pitching-control data. can experience hitting for various types of baseballs (B).

The display terminal 74 is installed in a position adjacent to the at-bat of the batting game room 910, but is installed in a position where the pitching device 73 or the player hits, the maximum collision with the baseball (B) does not occur, or a separate It is preferable to be mounted on a protective means.

The controller 71 manages and controls the operations of the components 72, 73, 74, 75, and 76 of the batting game room 910.

In addition, when the controller 71 receives the transmitted and received lidar signal information from the lidar sensors 72 of the batting game room 910, the received lidar signal information is transmitted in the same manner as the controller 51 of FIG. 10 described above. After scanning the multi-signal information to generate lidar images for each cube (C2), edge images for each cube (C2) and center lines for each cube (C2) are extracted.

In addition, when the edge image for each cube C2 is generated and the center line information for each cube C2 is extracted, the controller 71 utilizes these to determine the movement of objects (player, bat, and baseball B) and physical characteristics of the object The vector information is detected, and the detected object vector information is reflected in the baseball-simulation program so that the result of the player's hitting is reflected in the baseball-simulation.

In addition, when the object vector information is detected, the controller 71 generates comment request data matching the detected object vector information and baseball-simulation environment information, and then transfers the generated comment request data to the AI-based operation server 3. , and when the comment information is received from the AI-based operation server 3, the received comment information is transmitted to the display terminal 74 so that the display terminal 74 displays the comment information on the player's strike.

At this time, the environmental information includes the current score, current ball counting, running situation, defensive position, etc. in the baseball-simulation.

FIG. 16 is an exemplary view showing a game room for pitching in FIG. 14 .

The game room 920 for pitching in FIG. 16 is a game room for providing a pitcher-based experiential baseball game and baseball-simulation service to players and at the same time providing a teaching service for the player's pitching.

In addition, the following accessories are installed in the game room 920 for pitching.

As shown in FIG. 16, accessories of the pitching game room 920 include a controller 71 and lidar sensors 72 that perform the same operations as the batting game room 910 of FIG. 15 described above. , a display terminal 74, a screen 75, and an image output device 76.

That is, in the game room 920 for pitching, among the accessories of the game room 910 for batting, auxiliary equipment except for the pitching device 73 are equally installed.

In addition, pitcher-based baseball-simulation images, not batter-based baseball-simulation images, are displayed on the screen 75 of the pitching game room 920 .

FIG. 17 is an exemplary front view for explaining a cube of the game room used for object detection by the controller of FIG. 14 , and FIG. 18 is an exemplary side view of FIG. 17 .

When the controller 71 of the present invention receives lidar signal information transmitted and received from lidar sensors 72 in each game room, it scans the received lidar signal to obtain lidar images, but in FIGS. 17 and 18 As shown, a lidar image for each cube C2 is obtained by dividing the game room into a plurality of pieces.

At this time, the cubes (C2) mean segments (Segments) that divided the competition room in the vertical and horizontal directions.

In addition, when the lidar images for each cube C2 are generated by the lidar sensors 72, the controller 71 analyzes the generated lidar images for each cube C2 to determine the number of objects (player, bat, and baseball). Object vector information (movement, trajectory, physical characteristics, etc.) is detected, and in detail, 1) the player's movement (movement and speed of each body, weight movement, etc.) and 2) the movement of the bat (rotational trajectory, speed, Contact position of bat and baseball, etc.), and 3) trajectory (moving direction, speed, rotation, trajectory, etc.) of baseball B in the game room are detected.

FIG. 19 is a block diagram illustrating the controller of FIG. 14 .

As shown in FIG. 19, the controller 71 includes a control unit 710, a memory 711, a data transceiver 712, a baseball-simulation operation unit 713, an object vector information generation unit 715 for practice, and a comment It consists of an information request unit 716 and a display unit 717.

The control unit 710 is an operating system (OS) of the controller 71, and manages and controls the operations of the control targets 711, 712, 713, 715, 716, and 717.

In addition, when the control unit 710 receives lidar signal information from the lidar sensors 72 through the data transceiver 712, it inputs the received lidar signal information to the object vector information generator 715 for practice.

In addition, when the object vector information for practice is generated by the object vector information generation unit 715 for practice, the control unit 710 inputs the generated object vector information for practice to the baseball-simulation operation unit 713 and the comment information request unit 716. .

In addition, when comment request data is generated by the comment information request unit 716, the control unit 710 controls the data transceiver 712 to transmit the generated comment request data to the AI-based operation server 3.

In addition, when receiving comment information from the AI-based operation server 3 through the data transmission/reception unit 712, the control unit 710 transmits the received comment information to the display terminal 74 so as to transmit the received comment information to the corresponding player in the display terminal 74. To display comment information about pitching or batting.

In the memory 711, installation location information of each of the lidar sensors 72 and location information of each of the preset cubes C2 are preset and stored.

In addition, the memory 711 includes LiDAR images for each cube C2 generated by the object vector information generation unit 715 for practice in FIG. 20 described later, edge images for each cube C2, and center line for each cube C2. Information and object vector information for practice are temporarily stored.

In addition, a pre-made baseball-simulation program is stored in the memory 711 .

At this time, the baseball-simulation program proceeds with a baseball game based on a virtual stadium simulating a real stadium, but when object vector information for practice is input, it analyzes the input object vector information for practice using a preset calculation formula or calculation table to obtain 1 ) When a player is a batter, simulation result information including batting results (missing swing, single base hit, double hit, home run, etc.), landing point and movement information of baseball (B) due to hitting, movement information of other players, etc. Detects and reflects the detected simulation result information to the background image to generate a baseball-simulation image, 2) When the player is a pitcher, the physical characteristics of the pitched baseball (B), the pitched baseball (B) Detects simulation result information including location information of the strike-zone, batting results (miss swing, single, double hit, home run, foul, etc.), motion information of other players, etc., and sends the detected simulation result information to the background It is reflected on the image to create a baseball-simulation image.

The data transceiver 712 transmits and receives data with the AI-based operation server 3.

The baseball-simulation operation unit 713 interworks with the AI-based operation server 3 for every predetermined period, and performs an update of the baseball-simulation program.

At this time, the AI-based operation server 3 analyzes the data collected by the data collection systems 800 to generate ability value information for each player, and then performs baseball-simulation so that the generated ability value information for each player is reflected in the baseball-simulation program. By updating the program every cycle, the movement of the actual baseball player, the movement of the actual bat, and the trajectory of the actual baseball B are reflected in the baseball-simulation program, thereby maximizing the reality and realism of the baseball-simulation.

That is, the baseball-simulation operation unit 713 works with the AI-based operation server 3 for every period to perform an update of the baseball-simulation program.

In addition, the baseball-simulation operation unit 713 is executed when a baseball-simulation service is requested from a player in the game room, and operates a baseball-simulation game based on a pre-made baseball-simulation.

At this time, the baseball-simulation program of the baseball-simulation operation unit 713 proceeds with a baseball game based on a virtual stadium simulating a real stadium, and generates a baseball-simulation image in real time. At this time, the generated baseball-simulation image is transmitted to the image output device 76 through the data transmission/reception unit 712 under the control of the controller 710 and projected on the screen 75.

In addition, the baseball-simulation operating unit 713 provides a batter-based baseball-simulation image when a player is a 'batter', but throws a baseball B by linking the motion of the pitcher object in the image with the pitching device 73. make this happen

That is, the baseball-simulation operation unit 713, when the player is a 'batter', when the pitching object on the baseball-simulation is ready, the speed, rotation, pitch type, strike zone location, etc. for the next pitching from the baseball-simulation program is received, and pitching-control data is generated by matching them. At this time, the generated pitching-control data is transmitted to the pitching device 73 under the control of the controller 710, and the pitching device 73 throws the baseball B according to the received pitching-control data.

In addition, when the baseball-simulation operating unit 713 receives the object vector information for practice from the object vector information generation unit 715 for practice, it inputs the input object vector information for practice into a baseball-simulation program, and the baseball-simulation program is input for practice. After operating the baseball-simulation based on the object vector information, the pitching result or batting result according to the input object vector information for practice on the virtual baseball field is detected.

At this time, the baseball-simulation operation unit 713 extracts result information (a pitching result or a batting result) from the baseball-simulation, and then inputs the extracted result information to the comment information request unit 716.

In addition, the baseball-simulation operation unit 713 extracts environmental information including the current score, current ball counting, running situation, and defensive position from the baseball-simulation program, matches them to generate environmental information, and generates environmental information. to the comment information request unit 716.

FIG. 20 is a block diagram showing an embodiment of the object vector information generation unit for practice shown in FIG. 19 .

The first object vector information generator 715-1, which is an embodiment of the object vector information generator 715 for practice in FIG. , a processor for generating object vector information for at-bat including movement and physical characteristics of objects including bats and baseballs.

In addition, as shown in FIG. 20, the first object vector information generation unit 715-1 includes a lidar sensor input module 7151, a scanning module 7152, an edge image conversion module 7153 for each cube, and object recognition module 7154, object tracking module 7155, player movement information generation module 7156, bat movement information generation module 7157, baseball trajectory information generation module 7158, object vector information generation module for turn at bat 7159 made up of

The lidar signal input module 7151 receives lidar signal information transmitted and received by lidar sensors 72 in the corresponding game room.

The scanning module 7152 acquires a lidar image by scanning the lidar signal of the lidar sensor 72 input through the lidar signal input module 7151, and obtains the scanning information and each preset cube (C2 ), LIDAR images are generated for each cube (C2).

As described above in FIG. 11, the cube-specific edge image conversion module 7153 applies a Sobel filter to the cube-specific LiDAR image generated by the scanning module 7152 to convert the cube-specific LiDAR images into cube-specific edge images. convert each to

The object recognition module 7154 analyzes the edge image for each cube C2 converted by the edge image conversion module 7153 for each cube C2 to recognize objects (batter, baseball, and bat).

In addition, when the recognized object is the player P, the object recognition module 7154 calculates the center of the player object from the edge image for each cube converted by the edge image conversion module 7153 for each cube, as described above in FIG. 12. After extracting the lines, the positions of the head (A), hands (B, C), and feet (D, E) of the body are determined by utilizing the center line 431 of the extracted player object P.

The object tracking module 7155 tracks movements of objects (player, baseball, and bat) recognized by the object recognition module 7154.

At this time, the object tracking module 7155 tracks the movement of each body part of the batter player when the recognized object type is the batter player.

The player motion information generation module 7156 recognizes the movement of each body part (arm, hand, shoulder, foot, etc.) of the player that is recognized by the object recognition module 7154 and tracked by the object tracking module 7155. (moving direction, speed and rotational trajectory, etc.) are detected.

In addition, the player motion information generation module 7156 detects weight movement information, which is the central axis of the body, through movement information of the center line recognized by the object recognition module 7154.

That is, the player motion information generation module 7156 generates player motion information in which weight movement information is matched with the movement direction, speed, and rotational trajectory of each body part (arm, hand, shoulder, foot, etc.).

Figure 21 (a), (b) is an exemplary view for explaining the collision information of the bat and baseball of the present invention.

The bat movement information generation module 7157 detects the movement (moving direction, speed, rotational trajectory, etc.) of the bat recognized by the object recognition module 7154 and tracked by the object tracking module 7154.

In addition, the bat motion information generating module 7157 causes the surface of the bat 900 recognized by the object recognition module 7154 to collide with the baseball B, as shown in (a) and (b) of FIG. 21 . When the bat 900 and the baseball (B) collide, bat collision information (L) indicating the position and area on the bat 900 is detected.

In other words, the bat motion information generation module 7157 of the present invention generates bat motion information by matching 1) the bat collision information (L) with 2) the moving direction, speed, and rotational trajectory of the bat.

The baseball trajectory information generation module 7158 detects the location, movement speed, trajectory, firing angle, etc. of the baseball, recognized by the object recognition module 7154 and tracked by the object tracking module 7155, Then, baseball trajectory information is generated by matching the detected values.

At this time, as described above, in the present invention, since the trajectory of the baseball B is divided into cubes C2 for detection and tracking, the trajectory of the baseball B can be accurately and quickly tracked.

The object vector information generation module 7159 for a turn at bat generates player movement information generated by the player movement information generation module 7156, the bat movement information generation module 7157, and the baseball trajectory information generation module 7158, the bat movement information and The baseball trajectory information is matched to generate object vector information for at-bat.

At this time, the object vector information for a turn at bat generated by the object vector information generation module 7159 for a turn at bat is input to the baseball-simulation operation unit 713 and the comment information request unit 716 under the control of the control unit 710, and , The baseball-simulation program of the baseball-simulation operation unit 713 reflects the input object vector information for at-bat and operates baseball-simulation based on this, thereby maximizing the reality and reality of the batter player.

22 is a block diagram showing a second embodiment of the object vector information generation unit for practice shown in FIG. 19;

The second object vector information generation unit 715-2, which is the second embodiment of the object vector information generation unit 715 for practice shown in FIG. 22, is applied when the controller 71 is installed in the pitching competition room 920. It is a processor for generating object vector information for pitching including movement, trajectory and physical characteristics of objects including the player and baseball B.

In addition, as shown in FIG. 22, the second object vector information generation unit 715-2 has the same configuration as the first object vector information generation unit 715-1 of FIG. 20 described above (LIDAR signal input module 7151 ), a scanning module 7152, an edge image conversion module 7153 for each cube, an object recognition module 7154, and an object tracking module 7155.

In addition, the second object vector information generation unit 715-2 further includes a second player motion information generation module 7256, a second baseball trajectory information generation module 7258, and a pitching object vector information generation module 7259. include

The second player motion information generation module 7256 recognizes by the object recognition module 7154 and tracks each body part (arm, hand, shoulder, foot, etc.) of the pitcher player by the object tracking module 7155. ) detects the movement (direction of movement, speed and rotational trajectory, etc.).

In addition, the second player motion information generation module 7256 detects weight movement information, which is the central axis of the body, through movement information of the center line recognized by the object recognition module 7154.

That is, the second player motion information generation module 7256 generates player motion information in which the weight movement information and the movement direction, speed, and rotational trajectory of each body part (arm, hand, shoulder, foot, etc.) are matched.

The second baseball trajectory information generation module 7258 determines the location, movement speed, trajectory, and launching upon hitting of the baseball B, recognized by the object recognition module 7154 and tracked by the object tracking module 7155. angle, etc.

At this time, as described above, in the present invention, since the trajectory of the baseball B is divided into cubes C2 for detection and tracking, the trajectory of the baseball B can be accurately and quickly tracked.

In addition, the second baseball trajectory information generation module 7258 detects a position where the baseball B collides with the screen 75 (hereinafter referred to as screen-position information).

In addition, the second baseball trajectory information generation module 7258 matches the detected 1) position, moving speed, trajectory, firing angle, etc. of the baseball B at the time of hitting and 2) screen-location information to generate baseball trajectory information generate

The object vector information generation module 7259 for pitching matches the player motion information and the baseball trajectory information generated by the second player motion information generation module 7256 and the second baseball trajectory information generation module 7258 to obtain pitching information. Create object vector information.

At this time, the object vector information for pitching generated by the object vector information generation module 7259 for pitching is input to the baseball-simulation operation unit 713 and the comment information request unit 716 under the control of the control unit 710, while , The baseball-simulation program of the baseball-simulation operation unit 713 reflects the input object vector information for pitching, and operates baseball-simulation based on this, thereby maximizing the reality and reality of the pitcher player.

Returning to FIG. 19 and examining the comment information requesting unit 716, the comment information requesting unit 716 receives input from the first object vector information generating unit 715-1 or the second object vector information generating unit 715-2. received object vector information for practice (object vector information for at-bat or object vector information for pitching).

In addition, the comment information request unit 716 receives environmental information from the baseball-simulation operation unit 713 .

In addition, the comment information requesting unit 716 generates comment request data including the input environment information and object vector information (object vector information for a turn at bat or for pitching).

At this time, the comment request data generated by the comment information request unit 716 is transmitted to the AI-based operation server 3 through the data transmission/reception unit 712 under the control of the control unit 710.

The display unit 717 displays the comment information received from the AI-based operation server 3 on the display terminal 74 under the control of the control unit 710.

- AI-based operation server

23 is a block diagram showing the AI-based operation server of FIG.

As shown in FIG. 23, the AI-based operation server 3 includes a control unit 30, a database unit 31, a communication interface unit 32, an AI learning unit 33, a baseball-simulation management unit 35, It consists of an object vector information generation unit 36 for collection, an optimal pitching/hitting information generation unit 37, and a comment information generation unit 38.

The control unit 30 is the O.S (Operating System) of the AI-based operation server 3, and the control targets 31, 32, 33, 35, 36, 37, 38 Manage and control the operation.

In addition, when the control unit 30 receives transmission of edge images for each cube and center line information for each cube from the controller 51 of the data collection system 5 through the communication interface unit 32, it stores them in the database unit 31 and At the same time, the ability value information for each player is input to the generator 36.

In addition, when the ability value information for each player is generated by the ability value information generation unit 36 for each player, the control unit 30 stores the generated ability value information for each player in the database unit 31 .

In addition, the control unit 30 executes the AI learning unit 33 every preset first cycle T1.

In addition, the control unit 30 executes the baseball-simulation management unit 35 every second predetermined period (T2).

In addition, when the control unit 30 receives comment request data (object vector information for practice, environment information, and simulation-result information) from the controller 71 of the experiential baseball game operating system 7 through the communication interface unit 32, The received comment request data is input to the comment information generating unit 38 and stored in the database unit 31 at the same time, and when comment information is generated by the comment information generating unit 38, the generated comment information is stored in the controller ) to control the communication interface unit 32 to be transmitted.

In the database unit 31, the location and communication identification information of the controller 51 of each data collection system 5 are preset and stored.

In addition, in the database unit 31, the position of the controller of each experience-type baseball game operating system 7, communication identification information, and experience type information are preset and stored.

In addition, in the database unit 31, the position and identification information of each of the lidar sensors 53 installed in the baseball stadium 800 where each data collection system 5 is installed, and the cube C1 set in the baseball stadium 800 Location information of each of these is preset and stored.

In addition, in the database unit 31, the position and identification information of each of the lidar sensors 72 installed in each game room of the experience type baseball driving range 900 in which each experience type baseball game operating system 7 is installed, and the corresponding game room Location information of each of the cubes C2 set in is preset and stored.

In addition, the database unit 31 stores the optimal motion detection algorithm learned by the AI learning unit 33 .

Also, in the database unit 31, a baseball-simulation program that is pre-produced and optimized for every second period T2 by the baseball-simulation management unit 35 is stored.

In addition, the database unit 31 stores optimum pitching/hitting information generated by the optimal pitching/batting information generator 37 and comment information generated by the comment information generator 38 .

The communication interface unit 32 is connected to the data collection systems 5-1, ..., (5-N) and experience-type baseball game operating systems 7-1, ..., (7-N). send and receive data

Under the control of the control unit 30, the AI learning unit 33 is executed every preset first cycle T1 and learns an optimal motion detection algorithm.

At this time, the optimal motion detection algorithm utilizes the object vector information for practice, environment information, and simulation-result information of each game room collected during the first period (T1), 1) when a player is a batter, a baseball-simulated virtual baseball field Parameters for detecting the optimal hit (movement of each body, bat movement including bat collision position (L)) according to the environment (current score, current ball counting, base running situation, defensive position, etc.) of the batter player in Derives a set of values, and 2) when the player is a pitcher, detects the optimal pitching (movement of each body, trajectory information of the baseball (B)) according to the environment of the pitcher-player in the virtual baseball field of baseball-simulation Derive a set of parameter values for

The optimal motion detection algorithm learned by the AI learning unit 33 is used in the optimal pitching/batting information generator 37 .

The baseball-simulation management unit 35 is executed according to the control of the control unit 30 every second period T2, which is set in advance, and during the second period T2, the ability value information for each player generated by the generator 36 is generated. After collecting ability information, the baseball-simulation program is updated so that the collected ability information for each player is reflected in the baseball-simulation program.

In addition, the baseball-simulation management unit 35 interworks with the baseball-simulation operating unit 713 of the controller 71 of each experiential baseball game operating system 7 to update the baseball-simulation program stored in each controller 71. carry out

24 is an exemplary diagram illustrating a process of detecting motion information of a batter, a pitcher, a fielder, and a catcher from the point of time when a batter hits a ball in an actual baseball game.

In general, a baseball game consists of offense and defense, offense consists of hitting and stealing, hitting consists of full swing, bunt, cut-off, infield hit, outfield hit, etc., and defense consists of pitcher, catcher, It consists of infield defense, outfield defense, etc., and each position player, depending on their position, from the first time the bat swings, the first time the bat is made, or the first time the steal starts, that is, the event occurs from the first time Depending on the reaction, they have different reaction movements, and these reaction movements have characteristics that vary due to various variables such as physical ability, position, and individual judgment.

In view of the characteristics of baseball, the present invention is configured so that object detection and tracking are performed for each cube C1 of the baseball stadium 800, so that when an event occurs, an object (player or ball) is included from the time the event occurs. Respond to each player's event by utilizing the location information of each cube (C1) and the location coordinates within the cube (C1), and at the same time dividing and tracking the movement of these objects based on the cube (C1) according to the timeline. By accurately detecting movements according to the timeline, it is possible to precisely detect ability values such as reaction speed and movement speed of each player.

For example, as shown in FIG. ), a cube containing a ball, a cube containing a pitcher 220 (S21), and a cube containing a fielder 230 (S31) are detected. After that, the players 210, 220, and 230 have different reaction movements, as shown in (b) and (c) of FIG. 24 according to their positions. In the present invention, the game analysis algorithm After detecting the cubes including each object according to the timeline, the location of each object can be accurately tracked using the location information of each detected cube and the location coordinates (x, y, z) within the cube.

That is, according to the object detection and tracking based on the cubes C1 obtained by dividing the entire area of the baseball stadium 800 into a plurality of pieces, the ability value information generation unit 36 for each player of the present invention, when a batter hits, time The movements of the batter 210, the pitcher 220, and the fielder 230 along the line can be accurately recognized, for example, through the movement (speed and direction) of the fielder 220 after the hit is made. It is possible to estimate the fielder's ability value for the reaction speed for recognizing the ball's landing point, and through the movement of the pitcher 220, it is possible to estimate the defense pattern after the hitter's hit according to the current running state.

In addition, since object detection and tracking are performed based on the cubes (S) in which the entire stadium is divided into segments, the player-specific ability value information generation unit 36 eliminates unnecessary calculation processing, thereby significantly improving the processing speed. Precise tracking of objects is possible, and each player's reaction movement for a specific event can be accurately detected.

FIG. 25 is a block diagram showing a player-by-player ability value generation unit of FIG. 23 .

The ability value information generation unit 36 for each player in FIG. 25 analyzes the edge image for each cube C1 and the center line information for each cube C1 received from the controller 51 of the data collection system 5, and It is a processor for detecting capability level information.

In addition, as shown in FIG. 25, the ability value information generator 36 for each player includes a data input module 360, an analysis module 361, an object recognition/tracking module 362, a current state recognition module 363, Batter movement information generation module 364, pitcher movement information generation module 365, fielder movement information generation module 366, catcher movement information generation module 367, situational correspondence pattern information generation module 368, ability value for each player It consists of an information generating module 369.

The data input module 360 receives edge images for each cube C1 and center line information for each cube C1 received from the controller 51 of the data collection system 5 .

The analysis module 361 analyzes the edge images of each cube C1 and the center line information of each cube C1 input by the data input module 360 by utilizing the preset location information of each cube C1. .

The object recognition/tracking module 362 detects an object through the analysis data detected by the analysis module 361, recognizes the type of the detected object, and detects location information of the object at the same time. At this time, the object includes a player, a bat, and a baseball (B).

In addition, when the recognized object is a player, the object recognition/tracking module 362 utilizes the center line information as described in FIG. C), feet (D, E), etc. are detected.

At this time, in the drawings, for convenience of explanation, the body part is described as being composed of a head, hands, and feet as an example, but it is natural that the body part can be more subdivided and detected.

Also, the object recognition/tracking module 362 tracks the movement of the object through the location information of the recognized object.

The current state recognition module 363 is executed when the object recognized by the object recognition/tracking module 362 is a baseball player, and utilizes the object tracking information input from the object recognition/tracking module 362 to detect each player object. Be aware of the current state.

At this time, the current state can be detected as 'attack' and 'defense' in a wide range of categories, and the 'defense' category is 'batter', 'first baseman', 'catcher', 'infielder', 'outfielder; etc., and the 'attack' category can be classified into sub-components such as 'batter', 'full swing', 'single hit', 'cut off', 'bunt', The category of 'infielder' can be classified into sub-components such as 'preparation before pitching', 'ball catch', and 'throwing', and to a smaller extent, each player's own position and motion vector (moving direction and speed). ), position and motion vectors of players on the same team, running situation, ball position and motion vector.

In addition, the current state recognition module 363 inputs the object tracking information to the batter motion information generating module 364 when the current state of the player object is 1) a batter, and 2) when the player object is a pitcher, the object tracking information is converted into pitcher motion information. 3) When it is a fielder, the object tracking information is input to the fielder motion information generation module 366. 4) When it is a catcher, the object tracking information is sent to the catcher motion information generation module 367. 5) When it is a baseball (B), object tracking information is input to the baseball trajectory information generation module 368.

The batter's motion information generation module 364 generates motion information of the batter, which is motion information of an object recognized as a batter by the object recognition/tracking module 362.

At this time, the batter's movement information includes the type of batting (bunt, full swing, cut-off, etc.) information, the speed, trajectory and direction information of the batting, the type of ball, speed and direction information, and the batting result (hit, miss swing, home run) etc.) of information.

The pitcher motion information generating module 365 generates pitcher motion information, which is motion information of an object recognized as a pitcher in the object recognition/tracking module 362.

At this time, the pitcher motion information includes ball type, speed, trajectory and direction information, count result information, batter's batting result information, etc., and ball types include 'fastball', 'curve', 'sinker', and 'slider' ', 'change-up', 'fork', etc.

That is, since the pitcher's motion information is configured to generate information on all motions related to the actual pitching of the pitcher, reality and realism can be maximized in baseball-simulation.

The beast motion information generation module 366 generates beast motion information, which is motion information of an object recognized as a beast in the object recognition/tracking module 362.

At this time, the motion information of the fielder consists of information about the location, moving direction, jumping, diving, and speed of the fielder at the time of hitting, the ball catch result, and the trajectory and speed of the ball at the time of throwing.

The catcher motion information generating module 367 generates catcher motion information, which is motion information of an object recognized as a catcher by the object recognition/tracking module 362.

At this time, the catcher's movement information consists of pitcher's ball type, trajectory and speed information, count result information, catch information, ball speed and trajectory information at the time of throwing, etc.

The corresponding pattern information generation module 368 for each situation detects corresponding pattern information (position, movement direction and speed of the object) according to the game condition for each player.

At this time, the condition of the game consists of 'position', 'running condition', 'offensive condition', 'ball count condition', 'out count condition', and 'score condition'.

The ability value information generation module 369 for each player includes a batter movement information generation module 364, a pitcher movement information generation module 365, a fielder movement information generation module 366, a catcher movement information generation module 367, and a response pattern for each situation. Referring to the motion information and corresponding pattern information of each player generated by the information generation module 368, ability value information is generated/updated for each player.

At this time, the control unit 30 inputs the ability value information for each player generated/updated by the ability value information generation unit 36 for each player to the baseball-simulation management unit 35 .

FIG. 26 is a block diagram showing the optimal pitching/hitting information generation unit of FIG. 23 .

As shown in FIG. 26, the optimal pitching/hitting information generating unit 37 includes a data input module 371, an artificial intelligence-based analysis module 372, and an optimal pitching/hitting information generating module 3763.

The data input module 371 receives comment request data transmitted from the controller 71 of the experiential baseball game operating system 7 under the control of the control unit 30 .

At this time, the comment request data includes object vector information for practice, environment information, and result information.

The artificial intelligence-based analysis module 372 uses the input comment request data as input data of the optimal motion detection algorithm learned by the AI learning unit 33 to perform analysis.

The optimal pitching/hitting information generation module 373 includes 1) optimal movement information and weight movement information of each body of the pitcher player output from the optimal motion detection algorithm in the artificial intelligence-based analysis module 372, and the optimal pitch type. or 2) generate optimal pitching/batting information including optimal movement information and weight movement information of each body of the batter player, and optimal batting movement information.

FIG. 27 is a block diagram illustrating a comment information generation unit of FIG. 23 .

As shown in FIG. 27, the comment information generator 38 includes a data input module 381, a data extraction module 382 for each category, a data comparison module 383 for each category, a visualization module 384, and comment information. It consists of a generating module 385.

The data input module 381 includes the optimal pitching/hitting information generated by the optimal pitching/hitting information generation unit 37 and the comment request data (for practice) transmitted from the controller 71 of the experiential baseball game operating system 7. Object vector information, environment information, and result information) are input.

The category-specific data extraction module 382 extracts the input object vector information for practice and optimal pitching/batting information for each preset category. At this time, the category consists of 1) when the player is a pitcher, 'movements for each body of the player', 'weight transfer', and 'type of pitch', 2) when the player is a batter, 'movements for each body of the player', It consists of 'weight transfer' and 'bat movement'.

In general, in baseball, 1) in the case of a hitter, various batting results are achieved depending on various variables such as weight shift, movement of each body part, swing trajectory, swing strength, and swing trajectory, and 2) in the case of a pitcher, weight shift It is a very sensitive sport in which various results are achieved depending on various variables such as movement of each body part, ball type, ball trajectory, etc.

That is, the category-specific data extraction module 382 of the present invention takes into account the characteristics of baseball, the object vector information for practice related to the player's pitching/hitting, and the optimal pitching/batting information generated by the optimal pitching/hitting information generator 37. By extracting /batting information for each preset category, comments on the pitching/batting of the pitcher/batter player can be made accurately and quickly by comparing them in the future.

The category-by-category data comparison module 383 compares the data extracted by the category-by-category data extraction module 382, that is, compares the data of the same category in the object vector information for practice and the optimal pitching/hitting information, so that each category has a difference. detect data

For example, when the trajectory information of the arm body part of the object vector information for practice is 'x' and the trajectory information of the arm body part of the optimal pitching/hitting information is 'y', the data comparison module 383 for each category is ' 'l x - y l' may be detected as difference data by comparing data in the 'trajectory of arm body part' category.

The visualization module 384 visualizes difference data for each category detected by the data comparison module 383 for each category in graphics, pictures, text, symbols, and the like.

The comment information generation module 365 matches difference data for each category visualized by the visualization module 364 and simultaneously generates comment information including information such as problems and solutions.

At this time, the comment information generated by the comment information generation module 385 is transmitted to the controller 71 of the experience baseball game operating system 7 through the communication interface unit 32 under the control of the control unit 30 .

As such, the experiential baseball-simulation service system 1 using lidar sensors and artificial intelligence, which is an embodiment of the present invention, utilizes data collected through LiDAR sensors installed in actual baseball stadiums to determine the ability of each player. and at the same time reflect the generated ability of each player in the baseball-simulation, and operate the baseball-simulation based on the motion of the object detected through the lidar sensors installed in the experiential baseball driving range to increase reality and experience baseball game It is possible to increase the interest, realism, and participation of the player, and maximize the efficiency and effect of the player's baseball practice by providing comment information on the player's pitching or batting.

1: Experiential baseball-simulation service system using lidar sensor and artificial intelligence
3: AI-based operation server 5: Data collection system
7: Experiential baseball game operating system 10: Communication network
30: control unit 31: database unit
32: communication interface unit 33: AI learning unit
35: Baseball-simulation management unit 36: Ability information generation unit for each player
37: Optimal pitching/hitting information generation unit 38: Comment information generation unit
51: controller 53-1, ..., 53-N: lidar sensors
71: controller 72: lidar sensors
73: pitching device 74: display terminal
75: screen 76: video output device
511: control module 512: memory
513: data transmission/reception module 514: lidar signal scanning module
515: Edge image extraction module for each cube 516: Center line extraction module for each cube
517: data collection module 710: control unit
711: memory 712: data transmission and reception unit
713: baseball-simulation operation unit 715: object vector information generation unit for practice
716: Comment information request unit 717: Display unit
800: Baseball stadium 900: Experience type baseball practice field
910: game room for batting 920: game room for pitching

Claims (12)

Operation of an experience-type baseball driving range equipped with game rooms, which are either game rooms for at-bats that provide batter-based experiential baseball-simulation to players or pitching-type game rooms that provide pitcher-based experiential baseball-simulation to players. In the experiential baseball-simulation service system including an experiential baseball game operating system that manages:
The experiential baseball game operating system
a screen on which batter-based baseball-simulation images are output when installed in a batting game room, and pitcher-based baseball-simulation images are output when installed in a pitching game room;
LiDAR sensors that are installed at intervals in vertical and horizontal directions on the top and side of the game room for batting, transmit LiDAR signals to the interior space, and then receive reflected signals;
A controller that outputs a batter-based baseball-simulation image to the screen when installed in a batting game room, and outputs a pitcher-based baseball-simulation image to the screen when installed in a pitching game room;
It is installed in each at-bat game room and includes a pitching device that throws a baseball toward the at-bat through the throwing ball of the screen of the at-bat game room,
When the screen is installed in a game room for batting, a throwing ball is formed on one side,
The controller
When the segments obtained by dividing the interior space of the corresponding game room into a plurality of pieces are referred to as cubes C2, a memory storing location information of each cube C2;
After scanning the lidar signal transmitted and received by the lidar sensors to recognize and track the object of the player and baseball, an object vector information including player motion information and baseball trajectory information is generated. vector information generator;
Utilizing a pre-made baseball-simulation program, proceeding with a virtual baseball game based on a virtual stadium that simulates a real baseball stadium, using a preset calculation formula or calculation table to generate the object vector information for practice A baseball-simulation operation unit for generating a baseball-simulation image by deriving a result according to the object vector information for practice and matching it to a virtual baseball game, and outputting the generated baseball-simulation image to the screen,
When the controller is installed in the batting game room, the object vector information generation unit for practice
A scanning module for scanning lidar signals transmitted and received by the lidar sensors and generating lidar images for each cube by utilizing scanning information and preset location information for each cube (C2);
With respect to the LIDAR image for each cube C2 generated by the scanning module, a Sobel filter is applied to detect pixels having a pixel change rate equal to or greater than a threshold value, and an edge image for each cube C2 composed of edge lines is extracted. Edge image conversion module;
an object recognition module for recognizing objects including a batter player, a baseball and a bat by analyzing the edge image for each cube (C2) converted by the edge image conversion module for each cube;
an object tracking module for tracking motion of the object recognized by the object recognition module;
It is executed when the object tracked by the object tracking module is a batter player, and after detecting the movement of each body part of the batter player by utilizing the object tracking information detected by the object tracking module, the detected batter player's a player motion information generation module for generating player motion information including motions of each body part;
It is executed when the object tracked by the object tracking module is a bat, and by utilizing the object tracking information detected by the object tracking module, at least one of the recognized bat's moving direction, speed, and rotational trajectory, and the recognized bat's A bat motion information generation module for generating bat motion information by detecting and matching bat collision information (L) including the position and area on the bat where the bat and baseball collide;
It is executed when the object tracked by the object tracking module is a baseball, and by using the object tracking information detected by the object tracking module, at least one of the position, movement speed, trajectory, and launch angle at the time of hitting the recognized baseball is detected. a baseball trajectory information generation module for generating baseball trajectory information including one or more;
Object vector information for practice that generates object vector information for practice by matching the player motion information generation module, the bat motion information generation module, and the baseball trajectory information generation module to match the player movement information, bat movement information, and baseball trajectory information Experiential baseball-simulation service system, characterized in that it comprises a generating module. delete delete The method of claim 1, wherein when the controller is installed in the at-bat game room, the baseball-simulation operating unit
Experience-type baseball-simulation service system characterized in that the pitching device is controlled so that the pitching device is interlocked with the movement of the pitcher image of the baseball-simulation image so that the baseball (B) is thrown. The method of claim 4, wherein the baseball-simulation operating unit
After extracting at least one of the speed, rotation, pitch type, and strike zone location for the pitcher's next pitching from the baseball-simulation program, pitching-control data is generated by matching, and the pitching-control data generated is transmitted to the pitching device. do,
The pitching device
It includes a throwing control means for controlling the rotation, direction and speed of the baseball B during throwing, and when pitching-control data is received from the baseball-simulation operation unit, the pitching control means is controlled to receive the received pitching- An experiential baseball-simulation service system characterized in that the baseball (B) is thrown according to the control data. In claim 4 or 5, the controller
When the segments obtained by dividing the interior space of the game room into a plurality of pieces are called cubes C2, a memory for storing the location information of each cube C2 is further included,
When the controller is installed in the pitching competition room, the object vector information generation unit for practice
A scanning module for scanning lidar signals transmitted and received by the lidar sensors and generating lidar images for each cube by utilizing scanning information and preset location information for each cube (C2);
With respect to the LIDAR image of each cube C2 generated by the scanning module, a Sobel filter is applied to detect a pixel having a pixel change rate equal to or greater than a threshold value, and an edge image of each cube C2 composed of edge lines is extracted. Edge image conversion module;
an object recognition module for recognizing objects including a batter player, a baseball and a bat by analyzing the edge image for each cube (C2) converted by the edge image conversion module for each cube;
an object tracking module for tracking motion of the object recognized by the object recognition module;
It is executed when the object tracked by the object tracking module is a batter player, and after detecting the movement of each body part of the batter player by utilizing the object tracking information detected by the object tracking module, the detected batter player's a player motion information generation module for generating player motion information including motions of each body part;
It is executed when the object tracked by the object tracking module is a bat, and by utilizing the object tracking information detected by the object tracking module, at least one of the recognized bat's moving direction, speed, and rotational trajectory, and the recognized bat's A bat motion information generation module for generating bat motion information by detecting and matching bat collision information (L) including the position and area on the bat where the bat and the baseball collide;
Executed when the object tracked by the object tracking module is a baseball, and using the object tracking information detected by the object tracking module, the screen-position, which is the position where the recognized baseball collides with the screen, is detected, a baseball trajectory information generating module that generates baseball trajectory information including at least one of the detected screen-position, baseball location, moving speed, and trajectory;
The player motion information generation module and the object vector information generation module for practice that generates object vector information for practice by matching the player movement information and the baseball trajectory information generated by the baseball trajectory information generation module. Baseball-simulation service system. The method of claim 6, wherein the object recognition module
When the object recognized from the edge image of each cube (C2) is a player, after extracting the center line, which is a line connecting the median pixels, which are the intersections of the edge line of the player and horizontal lines formed evenly at regular intervals, the extracted center line to recognize the player's body,
The player motion information generating module
After detecting weight movement information, which is the central axis of the body, through the movement information of the center line recognized by the object recognition module, player movement information is generated to include the detected weight movement information. simulation service system. The method of claim 7, wherein the experiential baseball-simulation service system further includes an AI-based operating server and at least one data collection system built in a baseball stadium where an actual baseball game is held,
The data collection system
Second lidar sensors that are installed at intervals in horizontal and vertical directions on the side and top of the baseball stadium, transmit lidar signals toward the baseball stadium, and then receive reflected signals;
The positional information of each of the cubes C1, which are segments that divide the baseball field into a plurality of pieces, is preset and stored, and by scanning the lidar signal transmitted and received by the second lidar sensors, the radar signal for each cube C1 is stored. After generating this image, a Sobel filter is applied to the LiDAR image for each cube (C1) generated to extract an edge image for each cube (C2), and the extracted edge image for each cube (C1) is based on the AI Including a controller that transmits to the operation server,
The AI-based operating server
an ability value information generation unit for each player that recognizes and tracks an object by analyzing the edge image of each cube (C1) received from the controller, and then generates ability value information for each player using the recognized object tracking information;
It is executed every second period (T2), and after collecting the ability value information for each player generated by the ability value information generation unit for each player during the second cycle (T2), the collected ability value information for each player is baseball-simulated. Updating the baseball-simulation program to be reflected in the program, and interworking with the baseball-simulation operating unit of the controller, comprising a baseball-simulation management unit that transmits the updated baseball-simulation program system. 9. The method of claim 8, wherein the ability value information generation unit for each player
a data input module for receiving edge images for each cube (C1) transmitted from the controller of the data collection system;
an analysis module that analyzes the edge images of each cube C1 inputted by the data input module by utilizing preset location information of each cube C1;
After detecting an object including a player, a bat, and a baseball (B) through the analysis data detected by the analysis module, recognizing the type of the detected object and detecting the location information of the object at the same time, recognizing the object. an object recognition/tracking module that tracks the movement of an object;
a current state recognition module that is executed when the object recognized by the object recognition/tracking module is a baseball player and recognizes a current state of each player object by utilizing object tracking information input from the object recognition/tracking module;
It is executed when the object recognized by the object recognition/tracking module is a batter, and generates batter movement information including the movement of each body part of the batter object, batting type, speed, trajectory and direction of the bat, and batting result. Motion information generating module;
It is executed when the object recognized by the object recognition/tracking module is a pitcher, and pitcher movement information including movement of each body part of the pitcher object, pitch type, baseball speed, trajectory and direction, count result information, and batter betting result a pitcher motion information generating module that generates;
A fielder motion information generation module that is executed when the object recognized by the object recognition/tracking module is a pitcher and generates fielder motion information including movement of each body part of the fielder object, movement speed, ball catch result, and throwing accuracy. ;
a catcher motion information generation module that is executed when the object recognized by the object recognition/tracking module is a catcher and generates catcher motion information including movement of each body part of the catcher object, catch accuracy, and throw accuracy;
Referring to batter motion information, pitcher motion information, fielder motion information, and catcher motion information generated by the batter motion information generation module, the pitcher motion information generation module, the fielder motion information generation module, and the catcher motion information generation module, Including a capability information generation module for generating capability information for each player,
The current state applied to the current state recognition module is classified into 'attack' and 'defense', and the 'defense' category consists of sub-compositions of 'pitcher', 'first baseman', 'catcher', 'infielder', and 'outfielder'. The 'attack' category is classified into sub-components of 'batter', 'full swing', 'single hit', 'cut off', and 'bunt', and the category of 'infielder' is 'preparation before pitching'','ballcatch','throw' sub-components, characterized in that the experiential baseball-simulation service system. 10. The method of claim 9, wherein the baseball-simulation operating unit of the controller
Simultaneously extracting simulation-result information, which is a result of a player's pitching or batting, from the baseball-simulation program, and extracting environmental information including at least one of current ball counting, running situation, and defensive position,
The controller
a comment information request unit generating comment request data by matching the object vector information for practice generated by the object vector information generation unit for practice with the simulation-hitting result and environment information extracted by the baseball-simulation operation unit;
Experience-type baseball-simulation service system characterized in that it further comprises a control unit for transmitting the comment request data generated by the comment information request unit to the AI-based operation server. The method of claim 10, wherein the AI-based operation server
a database unit storing an optimal motion detection algorithm for outputting optimal pitching or batting information by using object vector information for practice of the comment request data received from the controller as input data;
It is executed every predetermined cycle (T1), and after collecting the comment request data transmitted from the controller during the first cycle (T1), the collected object vector information for practice of each game room, environment information, and simulation-result information are collected. 1) When a player is a batter, detecting the optimal hitting (movement of each body, bat movement including the bat collision position (L)) according to the environment of the batter player in the virtual baseball field of baseball-simulation 2) When a player is a pitcher, optimal pitching (movement of each body, baseball an AI learning unit for learning an optimal motion detection algorithm by deriving a set of parameter values for detecting a ball (B) (trajectory information);
It is executed when comment request data is received from the controller, and optimal pitching / hitting information for generating optimal pitching / hitting information by using the received comment request data as input data of the optimal motion detection algorithm learned by the AI learning unit Experience-type baseball-simulation service system, characterized in that it further comprises an optimal pitching / hitting information generator to generate. The method of claim 11, wherein the AI-based operation server further comprises a comment information generating unit,
The comment information generating unit
a data input module that receives the optimal pitching/hitting information generated by the optimal pitching/hitting information generator and the comment request data transmitted from the controller;
The object vector information for practice and the optimal pitching/hitting information input through the data input module are 1) when the player is a pitcher, at least one of 'movement for each body of the player', 'weight movement' and 'type of pitch' 2) a data extraction module for each category that extracts data for each preset category including at least one of 'motion of each body of the player', 'weight movement' and 'bat movement' when the player is a batter;
a data comparison module for each category that compares data of the same category of the object vector information for practice and optimal pitching/hitting information extracted by the data extraction module for each category and detects difference data for each category;
a visualization module that visualizes the difference data for each category detected by the data comparison module for each category using graphics, pictures, text, or symbols;
A comment information generation module for matching difference data for each category visualized by the visualization module, generating comment information including problems and solutions, and then transmitting the generated comment information to the controller;
The experiential baseball game operating system further includes a display terminal installed in each game room,
The controller
Experiential baseball-simulation service system characterized in that it further comprises a display unit for displaying the comment information received from the AI-based operation server on the display terminal.

Images