说明：  基于图模型的图像分割并行算法研究与实现,该方案通过网格划分来实现相似度矩阵的并行计算同时考虑到相似度矩阵的稀疏性和矩阵向量乘运算的内在并行性, 在该方案中本文设计并行算法来求解特征值问题在环境下的实验结果表明, 该并行方案是提高图模型分割方法实时性的有效途径.
(Image segmentation based on graph model of parallel algorithm and implementation of the program through the mesh to achieve similarity matrix of parallel computing the similarity matrix, taking into account the sparsity of matrix vector multiplication He' s inherent parallelism in the program in this article Design of parallel algorithms to solve the eigenvalue problem in the environment under the experimental results show that, 该 parallel programs is Tigao graph model segmentation in real time to Xing' s Youxiao way.)