结合Nystrm逼近的图半监督纹理图像分割

针对半监督学习算法在图像分割中的应用，提出了一种基于流形插值的半监督图像分割方法。该方法将分类问题看作一个流形上的函数的插值问题，通过优化某些系数来更好地拟合数据。该算法采用稀疏图可解决大规模矩阵特征值和特征向量的求解。但是，对于图像分割来说，构造稀疏图的运算时间较长，针对这一问题，提出采用Nystrm逼近方法来降低计算复杂度。合成纹理图像分割结果验证了该算法可获得良好的分割质量，结合Nystrm逼近方法在保证分割质量的前提下从很大程度上提高了计算效率。

Graph semi-supervised texture image segmentation combined with Nystr(o)m

To extend the application of semi-supervised learning in image segmentation,an image segmenta-tion method based on the manifold is proposed.This approach interprets the classification problem as a problem of interpolating a function on a manifold.Some coefficients are adjusted to provide the optimal fit to the data.The algorithm makes use of sparse adjacency matrix,which makes solving eigenvector problems for big matrix possible.However,it takes long time to construct the sparse adjacency matrix for image segmentation.To reduce computational complexity,an approach is proposed based on the Nystrom method,a numerical solution of eigenfunction problems.Experimental results of synthetic texture images segmentation indicate that the pro-posed method achieves good quality and using Nystr(o)m method improves the computational efficiency to a great degree.