低成本基于多幅图像的三维人体重建

提出一种基于多幅图像的低成本的重建目标三维人体模型的方法。首先建立每幅图像上特征点与原始模型特征点之间的对应关系，然后使用基于径向基函数方法对原始模型进行变形从而获得目标模型并对目标模型应用三种过滤器进行局部修复，最后使用视角无关的方法实现纹理映射和基于顶点混合的方法实现皮肤骨骼运动系统。实验结果表明该方法精度高，实用快捷且成本低廉。

Low-cost 3D human reconstruction from multi-image

A method for low-cost 3D specific human model reconstruction from multi-image is proposed. First, a series of feature points are matched on images and generic model respectively. Then a deformation from radial basic function is performed to obtain specific model. After that, three kinds of filters are used to refine the specific model. Finally the surface of model is texture mapped by viewpoint-independent algorithm, and it is skin-skeletoned by vertex blending algorithm as well. The running experiment results show that this method is accurate, easy to implement and low-cost.