论文部分内容阅读
将遗传算法与模糊C均值聚类算法(FCM算法)结合,并运用于磨粒图像目标提取。遗传FCM算法的基本思路是:首先对模糊聚类中心进行编码,然后依据FCM算法的目标函数建立适应度函数,在适当的交叉率和变异率下,最终实现了图像的有效分割。并考虑在一维分割特征向量情况下,通过引入直方图统计特性,实现了遗传FCM算法的快速运算。分割实验表明本文方法在一定程度上改善了标准FCM算法的性能,能有效地运用于智能铁谱分析系统中的磨粒目标自动提取。
The genetic algorithm and fuzzy C-means clustering algorithm (FCM algorithm) combined, and applied to the abrasive image target extraction. The basic idea of genetic FCM algorithm is as follows: Firstly, the fuzzy clustering center is coded, and then the fitness function is established according to the objective function of FCM algorithm. Finally, the effective segmentation of the image is achieved under the appropriate crossover rate and mutation rate. In the case of one-dimensional segmentation of eigenvectors, by introducing the statistical properties of histograms, the fast computation of genetic FCM algorithm is realized. Segmentation experiments show that this method improves the performance of standard FCM to some extent and can be effectively applied to the automatic extraction of abrasive particles in the intelligent iron spectrum analysis system.