摘要
对羽毛球运动员手臂击球图像轨迹跟踪,能够有效增强运动员击球质量。对击球手臂图像轨迹进行跟踪,需要计算击球手臂目标区域体态比和紧密度,过滤击球图像片段背景干扰物,完成对手臂击球图像轨迹的高效跟踪。传统方法结合自适应阈值分割方法,将击球手臂目标从背景中提取出来,但忽略了过滤掉击球图像片段背景干扰物。提出基于形态学算子的羽毛球运动员击球过程手臂击球图像轨迹跟踪方法。对连续两帧击球过程击球手臂图像序列进行差分计算,估计击球过程击球手臂差分图像高斯模型参数,提取运动员击球过程击球手臂运动目标的轮廓,计算击球手臂目标区域体态比和紧密度,过滤击球图像片段背景干扰物,构造运动目标全局匹配近似度函数,确定羽毛球运动员击球过程击球手臂运动轨迹。仿真结果表明,所提方法可有效跟踪击球过程中击球手臂目标,并生成连续击球手臂运动轨迹。
This article proposes a method for trajectory tracking of badminton player's arm stroke image in stroking process based on morphological operator. We carried out differential calculation for the image sequence of stroking arm in two consecutive frame stroke processes and estimated the parameters of Gaussian model of stroking arm differ- ential images in strokingproeess, then extracted the contour of motion ~arget of stroking arm in stroking process and calculated the posture ratio and compactness of target area of stroking arm. Moreover, the research filtered the frag- ments interferent of the image background and built a global matching approximation function of motion target. Final- ly, we determined the badminton player's trajectory of stroking arm in stroking process. Simulation result shows that the proposed method can effectively track the target of stroke arm in stroke process. It generates the trajectory of con- tinuous stroking arm.
作者
唐银春
蔺浩
TANG Yin - chun LIN Hao(Jinjiang College, Sichuan University, Meishan Sichuan 620860, China College of Physical Education, Chengdu University, Chengdu Sichuan 610106, China)
出处
《计算机仿真》
北大核心
2017年第10期229-232,256,共5页
Computer Simulation
关键词
羽毛球
手臂击球
轨迹跟踪
Badminton
Arm stroke
Trajectory tracking