基于图像视觉特征的体育训练错误动作捕捉

Sports Training Error Motion Capture Based on Image Visual Characteristics

由于教练无法对每名运动员的训练动作实时监督,一些错误动作轻则影响训练效果、重则导致运动员受伤。对此,提出一种基于图像视觉特征的体育训练错误动作捕捉。对训练图像进行预处理和判断当前帧是否存在训练动作;利用小波包分解将图像按照R、G、B划分得到三幅图像,分别提取每幅图像中第二层的16个信号特征,得到每幅图像的特征向量,将三幅图像特征向量归一化处理后,得到总的特征向量;通过与标准动作对比,判断其是否为错误动作实现捕捉。仿真分析中,所提方法能有效提取图像目标特征,且捕捉的错误动作无误判情况。

Because the coach can’t supervise every athlete’s training action in real time, some wrong actions may affect the training effect, or cause injury to the athlete. Therefore, a sports training error motion capture method based on image visual features is proposed. Preprocessing the training image and judging whether there is a training action in the current frame;The image is divided into three images according to R, G and B by wavelet packet decomposition. The 16 signal features of the second layer in each image are extracted respectively to obtain the feature vector of each image. After normalizing the feature vectors of the three images, the total feature vector is obtained;By comparing with the standard action, judge whether it is a wrong action to realize capture. In the simulation analysis, the proposed method can effectively extract the image target features, and the captured wrong actions have no misjudgment.