低维矩阵约束的非刚体特征点分类方法

Low dimension matrix constraint method for classifying non-rigid features

为了有效地对非刚体特征点进行分类,提出了一种低维子空间约束的非刚体特征点分类方法,该方法将所有非刚体特征点放入一个矩阵中,同时利用所有做刚体运动的特征点组成一个三维子空间的特性构造误差函数,误差最大的点可认为做非刚体运动,并逐一剔除差误,剩下的点即可认为做刚体运动.模拟实验和真实实验结果表明,该方法能够有效地实现非刚体特征点的分类.

To classify the features into the rigid and non‐rigid ones , a low dimension matrix constraint method for non‐rigid feature classification is proposed . All the non‐rigid features are put into a matrix and an error function is introduced based on the fact that the matrix which consists of rigid features spans a 3D subspace . The feature which has a max error is regarded as the non‐rigid one and moved from the matrix one by one . Experimental results with both simulated and real data show that the method can efficiently classify the features into two sets .