一种基于C-均值聚类的测量图像增强算法  

A Measurement Image Enhancement Algorithm Based on C-Means Clustering

作　　者：李晓冰[1] 马海潮[1] 高冰[1] 

机构地区：[1]92941部队

出　　处：《航空兵器》2008年第2期32-34,38,共4页Aero Weaponry

摘　　要：针对低质量红外测量图像人工搜索目标困难的问题,提出一种新的测量图像增强方法。该方法基于模糊C-均值聚类技术,对测量图像进行聚类分析,将各类分别赋予不同的灰度值,可增大目标边缘梯度。给出了此算法的具体原理和算法实现步骤。实验结果表明:此算法能够确保不损失目标信息的前提下,可最大限度地增强弱小目标图像。To solve the difficult problem of manually searching a target in an inferior infrared measurement image, a new measurement image enhancement algorithm is proposed which is based on fuzzy Cmeans clustering. According to this methed, a clustering analysis on the measurement image is carried out, and every cluster is allocatod to a different grey value so as to augment target edge grads. The algorithm principle and steps are described in detail. The experimental results show that the algorithm can greatly enhance small and weak target image in the state of ensuring target information not being damaged.

关 键 词：红外图像 图像增强 模糊算法 红外目标 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]