System for tracking, locating and calculating the position of an object in a game involving moving objects

1. A computerized system for automatic identification of various pitch-types of a ball itched by a pitcher in a game of baseball on a baseball field having a home plate, a pitcher's mound, a first base, a second base, and a third base, the system comprising:

at least two cameras wherein a primary camera is positioned above and within view of the home plate, and a side camera is positioned to the left or right of the home plate, wherein each camera captures a continuous video image; wherein the field of view of the primary camera is at least 120° and includes the positions of a batter (while on the home plate) and a catcher and a pitcher, and wherein the field of view of the side camera is at least 120° , and also includes the batter (while on the home plate) and the catcher and the pitcher;

wherein both cameras are in functional communication with, and transmit video data to a computer programmed with one of more backbone programs and an AI algorithm, receiving data from the backbone program, which is trained to classify a ball into various pitch types using the data provided by the one or more backbone programs,

wherein said one or more backbone programs defines and maps the 3-dimensional home plate space in real-time, wherein the home plate space is a 3-dimensional shape extrapolated from an N×N 2-dimensional grid centered on the home plate, and performs object detection, monitors and analyses the video images, and post-processes the video data from the cameras, and feeds it into the AI algorithm, and

wherein the one or more backbone programs detects and maps the ball in real-time, and calculates the speed of the ball and calculates and/or predicts the path of travel of the ball relative to the home plate space and predicting the entry and exit points of the ball relative to the home plate space, and produces data containing such information, and feeds said data to the AI algorithm which is trained to classify a ball into various pitch types using the data provided by the one or more backbone programs, and

wherein the pitch type data is transmitted from the computer to a screen viewable by a viewer.

2. The system of claim 1 wherein the N×N 2-dimensional grid is a 5×5 grid.

3. The system of claim 1 wherein the one or more backbone programs employs a bounding box and confidence prediction class probability map.

4. The system of claim 1 wherein the one or more backbone programs further identifies the ball in motion, continuously collects data about the location and speed of the ball, and provides this data to a strike prediction engine, wherein the data is transformed and analyzed to produce an output defining the pitch as a strike, a ball or a hit.

5. The system of claim 1 , further comprising a third camera is positioned to the right or left of home plate.

6. The system of claim 1 , further comprising a Doppler radar system poisoned at the location of the primary camera.

7. The system of claim 1 , wherein said baseball is marked to aid visual detection and assist disambiguation using a type of marker selected from the group consisting of color, infrared, ultraviolet and fluorescent markers.

8. The system of claim 1 wherein said at least two cameras start data collection on detection of a pitch trigger from hit detection engine.

9. The system of claim 2 wherein the 9 most central squares will be designated as a strike zone.

10. The system of claim 1 wherein the pitch type data is transmitted from the computer to a screen viewable by a viewer and wherein the pitch-type comprises a fast ball, a curve ball and a slider.

11. The system of claim 1 wherein the screen viewable by a viewer further displays information including the speed of the ball and whether the pitch was a strike, a ball or a hit.

12. The system of claim 1 wherein the screen viewable by a viewer further provides an option for instant replay and slow-motion video playback.

13. The system of claim 1 wherein the pitch type data is transmitted from the computer to a third-party system operated to provide streaming services to subscribers, and may optionally include data selected from the group consisting of: speed of the ball, pitch-type and data about whether the pitch was a strike or a ball.

14. The system of claim 1 wherein the backbone program performs the following steps:

frame acquisition, ball detection, ball tracking, determination of whether the pitch has been completed, speed estimation and path interpolation to predict the entry location and the exit location of the ball, and compiles data based on these characteristics, and where the AI algorithm is trained to perform, and performs pitch classification using the data provided by the one or more backbone programs.

15. The system of claim 1 wherein the primary camera is positioned semi-submerged in an enclosure, close to the pitcher's mound, but still be above and within view of the home plate.

16. A computerized system for automatic identification of various pitch-types pitched by a pitcher and/or classification of a ‘strike’ or a ‘ball’ or a ‘hit’ in a game of baseball on a baseball field having a home plate, a pitcher's mound, a first base, a second base, and a third base, the system comprising:

at least two cameras wherein a primary camera is positioned above and within view of the home plate, and a side camera is positioned to the left or right of the home plate, wherein each camera captures a continuous video image:

and wherein the field of view of the primary camera is at least 120° and includes the positions of a batter (while on the home plate) and a catcher and a pitcher, and wherein the field of view of the side camera is at least 120° , and also includes the batter (while on the home plate) and the catcher and the pitcher;

wherein both cameras are in functional communication with and transmit video images to a computer programmed with software running an AI-based object detection algorithm that monitors and analyses the video images comprising at least a Human Pose Estimation engine and a Strike Prediction Engine running a trained AI system, wherein the computer (i) uses Human Pose Estimation to ascribe facial and skeletal coordinates to the batter, (ii) ascribes coordinates to a home plate, a baseball mitt, and a baseball bat, if present in the frame, and (iii) integrates information about the coordinates of the batter and the coordinates of the home plate to define a real-time virtual three-dimensional volume (the strike zone), and further

(iv) identifies the ball in motion, continuously collects data about the location and speed of the ball, and provides this data to the Strike Prediction Engine,

wherein the data is transformed and analyzed to produce an output defining the pitch as a strike, a ball or a hit; and further,

wherein the AI system detects and maps the ball in real-time, and calculates the speed of the ball and calculates and/or predicts the path of travel of the ball relative to the home plate space and predicting the entry and exit points of the ball relative to the home plate space, and analyses the data so-produced by the AI-system to classify the pitch-type into at least a fast-ball, a slow-ball and a curve-ball.

17. The computerized system of claim 16 , further comprising a third camera is positioned to the right or the left of home plate.

18. The computerized system of claim 16 , wherein said baseball is marked to aid visual detection and assist disambiguation using a type of marker selected from the group consisting of color, infrared, ultraviolet and fluorescent markers.

19. The computerized system of claim 16 wherein the backbone program performs the following steps: frame acquisition, ball detection, ball tracking, determination of whether the pitch has been completed, speed estimation and path interpolation to predict the entry location and the exit location of the ball, and compiles data based on these characteristics, and where the AI algorithm is trained to perform, and does perform pitch classification, and further performs automatic identification of various pitch-types pitched by a pitcher and/or classification of a ‘strike’ or a ‘ball’ or a ‘hit’, using the data provided by the one or more backbone programs.

20. The computerized system of claim 16 wherein data produced by the AI-system is transmitted from the computer to a third-party system operated to provide streaming services to subscribers.