基于直方图移动平均的红外热图像自动模糊聚类分割方法  

A FUZZY CLUSTERING METHOD FOR AUTOMATIC SEGMENTATION OF INFRARED THERMAL IMAGE BASED ON HISTOGRAM MOVING SMOOTHING

作　　者：任彬[1] 汪炳权[1] 束学斌[1] 罗斌[1] 

机构地区：[1]安徽大学电子工程与信息科学系,合肥230039

出　　处：《模式识别与人工智能》1996年第3期278-282,共5页Pattern Recognition and Artificial Intelligence

摘　　要：根据红外热图像的灰度特性,提出了基于直方图移动平均的图像自动模糊聚类分割方法.对移动平均后的直方图采用峰值检测算法自动确定聚类数目.并对模糊C均值(FCM)聚类算法加以改进,使得算法速度大为提高,增加了FCM聚类算法的实用性.On accordance with the properties of infrared thermal image, A fuzzy clustering method for automatic segmentation of image based on histogram moving smoothing is discribes. The cluster number is obtained automatically by using the peak detection method after histogram smoothing and the algorithm of fuzzy C-mean (FCM) is modifyed in order to accelerate running speed, so. practical use of traditional method of FCM has been greatly increased.

关 键 词：图像处理 图像分割 模糊聚类分割法 红外图像 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]