基于混沌PSO或分解的二维最小误差阈值分割

Two-dimensional minimum error thresholding based on chaotic particle swarm optimization or decomposition

为了进一步提升现有的二维最小误差阈值分割快速递推算法的运行速度,提出分别基于混沌粒子群优化(PSO)和分解的2种二维最小误差阈值分割算法.第1种算法利用混沌粒子群优化算法搜寻二维最小误差法的最佳分割阈值,且在迭代过程的适应度函数计算中引入递推算法,大大减少了冗余计算;第2种算法将二维最小误差法的运算分解成一维最小误差法和一维最小类内对数方差法的运算,计算复杂度由二维递推算法的O(L2)进一步降为O(L).实验结果表明,提出的2种算法能够在分割效果达到或优于现有的二维最小误差阈值分割法的同时,大大加快运行速度.

A two-dimensional minimum error image thresholding method based on chaotic particle swarm optimization（PSO） or decomposition was proposed in order to further improve the computation speed of the fast recursive algorithm of existing two-dimensional minimum error thresholding method.The PSO-based method used the chaotic particle swarm optimization algorithm to find the optimal threshold of two-dimensional minimum error method.The recursive algorithm was adopted in the computation of fitness function in iterative procedure,and the redundancy computation was greatly reduced.For the second method,the computation of two-dimensional minimum error method was decomposed into the computation of one-dimensional minimum error thresholding method and one-dimensional minimum within-class logarithmic variance method,which made the computation complexity further reduced from O（L2） of the two-dimensional recursive algorithm to O（L）.Experimental results show that the two methods can greatly improve the running speed while the segmented results are as good as or better than the existing two-dimensional minimum error thresholding method.