基于背景差分法的篮球投篮轨迹预测模型构建  

Construction of Basketball Shot Trajectory Prediction Model Based on Background Difference Method

作　　者：吴晓军 WU Xiao-jun(Anhui Wenda College of Information Engineering,Hefei,Anhui 231201)

机构地区：[1]安徽文达信息工程学院,安徽合肥231201

出　　处：《河北北方学院学报（自然科学版）》2024年第5期27-33,共7页Journal of Hebei North University:Natural Science Edition

基　　金：“健康中国”背景下高校校园足球发展对策研究(项目编号XSK2022A07)。

摘　　要：为实现动态复杂环境下篮球投篮轨迹的预测,为篮球训练提供指导依据,构建基于背景差分法的篮球投篮轨迹预测模型。该模型基于混合高斯背景差分方法,构建复杂动态环境背景建模,为保证建模效果,引入自适应学习率进行模型优化,并删除背景模型中冗余高斯分量,获取图像中的前景目标图像;基于最大类间方差的阈值分割方法,二值化处理前景目标图像后,采用区域线性增长方法和篮球运动员投篮运动轨迹图形融合模型,提取前景目标图像中的投篮运动特征,实现篮球投篮轨迹预测。实验表明:该模型具有良好的篮球投篮运动图像处理效果,篮球投篮轨迹预测的可决系数指标结果均在0.947以上,预测结果与实际轨迹结果之间吻合程度较好,满足应用需求。In order to realize the prediction of basketball shot trajectory in dynamic complex environment and provide guidance basis for basketball training,a basketball shot trajectory prediction model based on background difference method was constructed.The model is based on the mixed Gaussian background difference method to build the complex dynamic environment background modeling.In order to ensure the modeling effect,the adaptive learning rate is introduced to optimize the model,and the redundant Gaussian component in the background model is deleted,so as to obtain the foreground target image in the image.Based on the threshold segmentation method of maximum inter-class variance,after binarization of foreground target image,the linear growth method of region and the fusion model of basketball players'shooting trajectory were used to extract the shooting motion features in the foreground target image,and the basketball shooting trajectory prediction was realized.The experimental results show that the model has a good effect on the image processing of basketball shooting motion,and the results of the determination coefficient of the prediction of basketball shooting trajectory are all above 0.947.The predicted results are in good agreement with the actual trajectory results,which meets the application requirements.

关 键 词：背景差分法 篮球投篮轨迹 预测模型 混合高斯方法 最大类间方差 冗余分量 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]