基于量子进化规划核聚类算法的图像分割

基于量子计算的并行性、进化计算简单、通用性好等优点,采用量子编码构造进化算法的染色体种群,再将二者引入到核聚类中来,提出了一种基于量子进化规划的核聚类算法.该算法充分利用了量子态的叠加性以及量子比特的概率表示,能够表示出许多可能的线性叠加状态,具有更好的种群多样性,因此将其用于解决核聚类算法中目标函数的优化问题,可以有效克服传统进化算法收敛速度慢以及早熟等问题.对Brodatz纹理图像及SAR图像进行分割,仿真实验结果表明该算法可以较好地改善图像分割效果.

Image Segmentation of Quantum Evolutionary Programming-based Kernel Clustering Algorithm

Quantum computation was introduced into fuzzy clustering based on its parallel character,the simplism and universality.So a new kernel clustering algorithm based quantum evolutionary programming is put forward.The method simulates quantum collapse and qubit probability characterization to describe many possible linear splice states.Compared with the traditional evolutionary algorithm,it has better population diversity.The algorithm is used to optimize the kernel fuzzy clustering and it can converge faster a...