基于模型的弹性连接体三维运动估计

Model-based 3D Motion Estimation of Elastic Articulated Objects

为解决非刚体运动分析中模型精度和计算复杂度之间的平衡问题,该文提出了一种基于圆锥曲线模型的弹性连接体(人体)三维运动估计方法。建立了人体三维模型,根据双目图像序列进行基于该模型的人肢体三维变形和运动参数估计,采用圆锥曲线三维模型及其在图像平面上投影的联系方程估计三维运动参数。人手臂运动估计实验结果表明该运动估计算法能够正确估计人体运动参数。

To offer a good tradeoff between computational complexity and model precision in non-rigid motion analysis,a method of 3D motion estimation for elastic articulated objects（human bodies） based on the conic curves is presented.A human body 3D model is established.The 3D deformation and motion parameters based on the proposed model are estimated by a sequence of stereo images.The 3D motion parameters are estimated based on the 3D model of conic curves and the corresponding equations of its two projections in image plane.The experimental results of human arm motion estimation demonstrate that the model-based estimation method can estimate the human motion parameters properly.