CN115138052A

CN115138052A - Basketball tactics analysis method based on ball and player tracks

Info

Publication number: CN115138052A
Application number: CN202210798893.3A
Authority: CN
Inventors: 蒋现新
Original assignee: Capital University of Physical Education and Sports
Current assignee: Capital University of Physical Education and Sports
Priority date: 2022-07-11
Filing date: 2022-07-11
Publication date: 2022-10-04

Abstract

The invention discloses a basketball tactic analysis method based on the tracks of a ball and players, which utilizes the track information of the basketball and the players on a court, carries out grid processing on the tracks, matches the tracks with a standard tactic library through a pattern recognition algorithm, recognizes tactic arrangements of the players and the players in real time, assists a coach in real-time decision-making, and improves the automation and the intellectualization of the basketball tactic analysis.

Description

Basketball tactics analysis method based on ball and player tracks

Technical Field

Sports analysis, basketball training

Background

In basketball preparation and training, tactical information needs to be acquired through a large amount of video analysis, which is a time-consuming and labor-consuming activity and is influenced by eye fatigue and subjective states. The invention obtains the movement track of the player through the IMU and obtains the movement track of the basketball through the video analysis of the top view angle. And the basketball movement track and the movement track of the players are processed and matched with the track or the rule of the tactical library, so that the tactical arrangement of the opposite attacking parties is automatically identified.

Disclosure of Invention

The method comprises four parts of basketball movement track acquisition, player movement track acquisition, track processing and tactical mode identification.

1) The basketball track is obtained mainly through top-view camera video analysis, and the movement track of the basketball is obtained through recognition of the basketball target and calculation by means of target detection calculation of deep learning. Because the IMU has a certain volume and weight, if the IMU is prevented from being inside a basketball, it may have an impact on the play or feel and feel of the basketball. Therefore, the invention does not adopt the IMU method to acquire the track of the basketball, but adopts a non-contact visual scheme to acquire the movement track of the basketball.

2) The movement locus of each player is mainly obtained by placing a wireless IMU on each player, and the IMU outputs XYZ space coordinates of the position in real time and then sends the XYZ space coordinates to an acquisition program in a wireless communication mode.

3) The track processing is mainly standardized, and the standard basketball court is 28 meters long and 15 meters wide. First, divide basketball court into 28 × 15=420 grids, each grid being 1 square meter. First, whenever a trajectory falls on a certain grid, the coordinates of the trajectory are changed to those of the grid. Second, if the straight-line distance of two players is less than 0.5 meters, they are considered to fall into the same grid, the selection of which is determined by the grid in which the midpoint of the distance of the two players is located. After standardization processing, the basketball track and the player track are changed into integer coordinates, and the range is limited, so that the space complexity is greatly reduced, and the tactical mode identification in the next step is facilitated.

4) The tactical pattern recognition is mainly to match the basketball track/player track with the tracks of the standard library so as to recognize the tactical pattern adopted by the basketball team.

Principle of operation of the system

1. Before the system starts working, a wireless 9-axis IMU is placed on each player, typically at the waist. So as to obtain the three-dimensional coordinates of the real-time position of the player in real time. Due to the drift phenomenon of the IMU, a top-view image segmentation mode is needed to help improve the position accuracy. Each IMU needs to have a unique ID, by which to distinguish which player is, and to distinguish the player's motion trajectory.

2. The picture of nobody on the basketball court is shot in advance as the background picture through the top view, so that the background separation is carried out on the top view picture, and the positions of the basketball and the players in the basketball court are identified.

3. And starting to work, acquiring the position information of each player on the basketball court in real time through the IMU, and identifying the movement track of the basketball through a target identification algorithm.

4. The basketball track and player track are processed into grids, the standard basketball court is 28 meters and 15 meters wide, and the standard basketball court is divided into 28 × 15=420 grids. Each grid is 1 square meter. First, whenever a trajectory falls on a certain grid, the coordinates of the trajectory are changed to those of the grid. Second, if two players are less than 0.5 meters in straight line distance, they are considered to fall into the same grid, the selection of which is determined by the grid in which the midpoint of the distance between the two players is located.

5. Matching the gridded basketball track and the team member track with the track of the standard tactical library through a pattern recognition algorithm, thereby recognizing the tactical patterns of the two parties. The standard tactical library includes trajectory information, dwell time information, and rules information.

6. Standard tactical libraries require a professional basketball coach or professional basketball analyst to define the trajectory of the basketball, the trajectory of the team members and the rules.

Technical requirements

1. And (4) capturing the trajectory of the team member in real time by adopting a wireless IMU.

2. The basketball track is identified from the top view direction, and the time continuity principle needs to be adopted in consideration of the shielding condition, and the track of the basketball is compensated in an interpolation mode when the basketball is shielded.

3. The basketball track and the player track need to be processed in a grid mode, and therefore the time and complexity of searching and matching are reduced.

4. A professional basketball coach or technical analyst is required to assist in building the standard tactical library trajectories and rules.

Drawings

Fig. 1 is a functional diagram of the whole system of the basketball tactical analysis method based on the track of the ball and the player, and describes the key components, main principles and main methods of the system.

The technical scheme has the advantages

The basketball tactical analysis method based on the ball and the player tracks, disclosed by the invention, has the advantages that the basketball tactical analysis is automated and intelligentized, the labor cost and the analysis practice are greatly reduced, and the accuracy of the reliability of the analysis result is improved.

Detailed Description

1) The track information of the player is acquired by the wireless IMU, a Bluetooth or WiFi wireless communication mode can be adopted, and the IMU needs to adopt 9 axes so as to acquire the XYZ space coordinates of the IMU through geomagnetic information

2) Background separation and image segmentation are carried out through top-view pictures, the track information of the basketball is obtained, meanwhile, the estimation information of basketball players is corrected, and the problem of integral drift of the IMU is solved.

3) The basketball track and the player track are processed in a grid mode, so that the complexity of the algorithm is reduced, and the real-time performance is improved.

4) And establishing a tactical standard library according to tactical information provided by a professional basketball coach or a technical analyst.

Claims

1. A basketball tactic analysis method based on the tracks of a ball and players is characterized in that the movement tracks of a basketball and players are obtained through an IMU and a top view camera, then the tracks are subjected to grid processing, and matching processing is carried out on the tracks and a tactic standard library, so that tactical information of both basketball attacking and defending parties is obtained in real time.

2. The data acquisition method of claim 1, wherein the trajectory is subjected to a gridding process.

3. The data acquisition method of claim 1, wherein integral drift information of the IMU is corrected by a top view camera.

4. The data acquisition method of claim 1, wherein tactical information of both basketball defenders and warfare partners is output in real time through tactical library matching.