摘要
针对篮球投篮角度人工矫正可参考性较差的问题,提出基于三维视觉的篮球投篮角度矫正方法。首先采集篮球投篮的三维视觉图像信息,采用小波多尺度分解方法对采集的图像进行降噪预处理;然后采用边缘轮廓特征提取方法提取图像中篮球投篮动作的形体轮廓特征,输入到形体矫正专家系统中进行视觉分析和矫正;最后进行仿真测试分析,结果表明,该方法的篮球投篮动作图像分析性能好,可以实现投篮角度准确实时矫正,指导篮球训练。
Since the manual correction for the basketball shooting angle has poor reference, a 3D vision based correction method for basketball shooting angle is proposed. The 3D visual image information of the basketball shooting is acquired. The wavelet multi-scale decomposition method is used to preprocess the acquired image for noise reduction. The edge contour feature extraction method is adopted to extract the body contour feature of the basketball shooting action in the image, and then the ex- tracted feature is input into the body correction expert system for visual analysis and correction. The method was performed with simulation test and analysis. The simulation and analysis results show that the method has good image analysis performance for the basketball shooting action, can accurately correct the shooting angle in real time and guide the basketball training.
出处
《现代电子技术》
北大核心
2017年第5期45-48,共4页
Modern Electronics Technique
基金
河南省软科学研究计划项目(豫科软(2015)067号
豫科软(2014)686号)
关键词
三维视觉
篮球投篮
角度矫正
图像分析
3D vision
basketball shooting
angle correction
image analysis
