结合颜色和纹理特征的岩心荧光图像含油级别判定  

Determination of the Oil-bearingGrade of Core Fluorescent Image byCombiningColor and Texture Features

作　　者：曹磊[1] 卿粼波[1] 滕奇志[1] 何海波 

机构地区：[1]四川大学电子信息学院,成都610065 [2]成都西图科技有限公司,成都610065

出　　处：《计算机与数字工程》2016年第2期311-316,共6页Computer & Digital Engineering

基　　金：国家自然科学基金(编号:61372174)资助

摘　　要：判定岩心荧光图像的含油级别对石油勘探研究具有重要的实际应用价值。不同的含油级别表现为一定的颜色和纹理特征。通过选定含油级别示例图像作为模板,结合颜色和纹理特征计算待判定图像与模板图像间相似度,进而判定荧光图像的含油级别。首先,根据主要颜色非均匀量化HSV颜色直方图提取荧光图像的颜色特征。然后,利用LBP算子生成LBP图像,并对LBP图像求取GLCM特征值,来描述荧光图像的纹理特征。最后,使用巴氏距离分别计算待判定图像的颜色相似度和纹理相似度,并多特征融合得出综合相似度。实验结果表明,结合颜色和纹理特征能够提高岩心荧光图像含油级别判定的准确性。It is of practical significance for petroleum exploration to determine the oil-bearing grade of core fluorescence image.Oil-bearing grades of core have different color and texture features.By selecting sample image of oil-bearing grade as a template,image similarity is computed and then oil-bearing grade of the fluorescent image is determined by combining color and texture features.Firstly,according to the major colors,non-uniform quantization of HSV color histogram is done.Secondly,LBP code is used to generate the LBP image and obtain the GLCM characteristic values of the LBP image to describe the texture feature of image.Finally,the color similarity and texture similarity of fluorescent image are calculated respectively using the Bhattacharyya distance,and comprehensive similarity of multi-feature fusion.Experimental results show that combining color and texture feature can improve the accuracy of the oil-bearing grade of core fluorescence image.

关 键 词：相似度 颜色直方图 LBP GLCM 巴氏距离 多特征融合 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]