A badminton trajectory tracking method, system, medium and equipment

Abstract

The application discloses a badminton trajectory tracking method, which includes: using the badminton model to detect the shuttlecock in the previous frame of the current frame of the badminton movement video to obtain the center of gravity of the badminton; extracting the motion foreground of the current frame; wherein, the motion foreground Including the badminton movement foreground and the interference object movement foreground; calculate the distance between the center of gravity and the center of gravity of each movement foreground, and sort the distances according to the size of the value to obtain the movement distance group; according to the preset filter conditions, the values ​​in the movement distance group Perform screening; mark the center of gravity of the motion foreground corresponding to the filtered value as the badminton position point of the current frame. Through the method of the invention, the tracking speed of the badminton trajectory can be improved and the frame loss phenomenon can be reduced when the shuttlecock trajectory is tracked. Correspondingly, the present invention also discloses a badminton motion trajectory tracking system, medium and equipment, which also have the above beneficial effects.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a method, system, medium and device for tracking a badminton movement track. Background technique [0002] In the process of badminton competition, the movement trajectory of the badminton is generally collected in real time through the camera, and then the running state of the badminton is analyzed by the relevant software system, so as to provide the coaches and players with corresponding reference. At present, there are two commonly used methods for tracking the trajectory of badminton. One is the TLD (tracking-learning-detection) target tracking algorithm, which detects and tracks moving objects based on the color histogram of the image. The real-time learning of the target object makes the algorithm run slowly and cannot track fast-moving objects such as badminton, and the running effect of this algorithm is easily affected by the background color. The other is the kcf (kern...

AI Technical Summary

Problems solved by technology

At present, there are two commonly used methods for tracking the badminton trajectory. One is the TLD (tracking-learning-detection) target tracking algorithm, which detects and tracks moving objects based on the color histogram of the image, but the algorithm The target object will be learned in real time, so that the running speed of the algorithm is relatively slow, and it cannot quickly track fast-moving objects such as badminton, and the running effect of this algorithm is easily affected by the background color

The other is the kcf (kernelized correlation filters) tracking algorithm, but when this algorithm tracks high-speed motion and small moving targets, it will lose frames in the process of tracking moving objects

Images