基于纹理特征的能见度测量时段区分方法  

Method of Distinguishing Visibility Measurement Periods Based on Texture Feature

作　　者：杨权[1] 黄山[1] 张洪斌[1] 汪勤[1] 张建军[1] 

机构地区：[1]四川大学电气信息学院,四川成都610065

出　　处：《现代电子技术》2008年第23期192-194,共3页Modern Electronics Technique

基　　金：四川省重点技术创新项目计划项目资助(07XM020)

摘　　要：为了进行能见度的数字摄像法测量,提出了基于纹理特征的能见度昼夜时段区分方法。从直方图分析中得到的纹理参数,包括灰度均值、方差、熵和扭曲度,均能反映出两类图像所特有的纹理特征。通过对纹理参数的统计分析,能够准确得到能见度测量的白天和夜间时段。将该方法用软件编程实现,表明该方法准确、可靠,为数字摄像法测量能见度的进一步研究奠定了基础。To measure the visibility by means of the digital photography visiometer system,a method of distinguishing visibility measurement periods of daytime and nighttime based on texture featurer is proposed. Parameters including gray mean, variance, entropy and skewness gotten from histogram could reflect the all texture features of images of daytime and nighttime. The daytime and nighttime can be obtained rightly by counting and analysing the texture parameters. The method which is realized by software programming show that it is accurate and reliable, and could lay a foundation for the further study of measuring the visibility based on the digital photography visiomter system.

关 键 词：纹理分析 能见度 直方图 数字摄像 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]