CAJ | 학술논문

提出一种不定形状基无差别约束中人体运动图像重构防失控算法，把运动员运动姿态的坐标看作一个随机过程，通过计算运动员的形状基数量，对无约束下的形状基进行分类，针对不同的形状基区域确定运动状态下形状基矩阵，运用分区约束结果指导不同重构矩阵的秩计算，对形状基数量进行可变性的计算，防止大区域下的重构失控。以不同的运动图像为实验对象，结果表明，该方法可以对运动员的三维运动参数重构结果进行优化，降低重构参数失控风险。
제출일충불정형상기무차별약속중인체운동도상중구방실공산법，파운동원운동자태적좌표간작일개수궤과정，통과계산운동원적형상기수량，대무약속하적형상기진행분류，침대불동적형상기구역학정운동상태하형상기구진，운용분구약속결과지도불동중구구진적질계산，대형상기수량진행가변성적계산，방지대구역하적중구실공。이불동적운동도상위실험대상，결과표명，해방법가이대운동원적삼유운동삼수중구결과진행우화，강저중구삼수실공풍험。
Put forward a kind of uncertain shape base indiscriminate constraints of human motion image reconstruction blowout out of control algorithm, the coordinates of the athletes of motion can be viewed as a random process, by calculating the shape of the base number of athletes, the shape of the base are classified under the unconstrained, according to different shape of the base area determine the motion state of form base matrix, the use of zoning constraints results to guide different reconstruction matrix rank of calculation, the shape of the base number of the calculation of variability, prevent large area under refactoring is out of control. In different motion image as experiment object, the results show that the method can be 3 d motion parameters of some athletes based on optimized results, reduce the risk of reconstruction parameters is out of control.