带有收敛因子的PSO-Fisher图像分割的方法

The Convergence Factorial Thresholding Image Segment Based on PSO-Fisher Algorithm

在大多数的实际应用中,仅仅将图像分成目标和背景两种类型是远远不够的。然而要实现三类、四类以致多类的分割,就必须研究多阈值的分割方法。以研究Fisher单阈值分割法为基础,利用带有收敛因子的PSO算法将Fisher评价函数来进行两阈值分割,最终推广到多阈值的基于Fisher评价函数的图像分割。并通过定误差图像分割实验证明了方法的有效性,保证了算法的收敛。同时证明了选择一个合适的粒子群规模可以使得带有收敛因子的PSO-Fisher算法在尽量少的迭代次数下提高算法的成功率。

In most application,it is not enough to classify imagine into goal and background.It is necessary to study multi-thresholding image segment in order to realize three,four and most type image segment.Two-thresholding and multi-thresholding image segment based on Fisher single-thresholding image segment were studied.The convergence factorial particle swarm optimization（PSO）-Fisher algorithm could assure convergence.The validity of Fisher segmentation method was indicated by fixed error image segmentation,and the convergence of the algorithm was ensured.The segmentation experimental results show that select an appropriate particle size group with a convergence of factors can make PSO-Fisher algorithm to minimize the number of iterations to improve the algorithm＇s success rate.