基于模糊C-均值聚类的靶场光测胶片成像质量检测方法  

Image Inspection Method of Optical Target Measurement Film at Range Based on Fuzzy Cluster Means

作　　者：王玉梅[1] 

机构地区：[1]92941部队96分队

出　　处：《战术导弹技术》2009年第4期17-20,25,共5页Tactical Missile Technology

摘　　要：针对光测胶片传统检测方式存在效率低，检测误差大，检测点的选择对检测结果影响较大等问题，根据模糊C-均值聚类的原理，提出了一种基于像素灰度聚类的胶片成像质量检测方法．利用扫描仪将光楔胶片进行数字化，并对其进行聚类分析，有效地消除了无信息图像区域，客观地反映了光测胶片的成像质量．试验证明，可提高效率3—5倍，并提高检测的准确性．Aimed at the fact that traditional method is characterized by low efficiency and high error, and the inspection result is affected by measurement spot, a method to inspect film image quality based on pixel gray cluster is presented according to the principle of fuzzy cluster means. Scanner is used to digitalize the film, and cluster analysis is carried out to eliminate the image without information and evaluate the image. It is proven that the efficiency and accuracy can be improved greatly.

关 键 词：光测胶片 检测 模糊C-均值聚类 灰度 直方图 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] O235[自动化与计算机技术—计算机科学与技术]