Capturing and Analyzing Volleyball Player Training Trajectory Data Based on Mean Shift Algorithm

To address the challenges of complex backgrounds and incomplete trajectory capture due to the fast movement of the target, this study proposes a data capture method for volleyball player training trajectories based on the mean shift algorithm.The human body model is considered a skeleton model with 51 degrees of freedom and 16 joints to digitize the training trajectory data, and dimensionality reduction is applied to reduce computational complexity.To reduce the dependency of the mean shift algorithm on environmental parameters, a probability density function from the gradient iterative estimation algorithm is selected, and the target's color information is used as a feature to complete the trajectory data capture.The experiment demonstrates that the method can capture the motion of each athlete's joint, achieving more accurate training trajectory data capture without depending on relevant parameters.