A Novel Texture Extraction Method for the Sedimentary Structures Classification of Petroleum Imaging Logging

摘要：

　　The technology for reservoir structure identification has become a challenging problem in the field of imaging logging technology.Because of the huge amount of information and a wide variety,it causes experts with low efficiency on the interpretation of reservoir evaluation and the performance depends highly on the individual experience(including cognitive level,visual decision,etc.).We proposed a new method for texture feature extraction based on macro and micro features.About 3320 imaging logging datasets are fed to support vector machine(SVM)to validate the gains of new method.As a result,the new proposed method achieved an Area Under roc Curve(AUC)value of 0.94.

关键词： Imaging logging Texture features Support Vector Machine (SVM) Area Under roc Curve (AUC)

作者: Haoqi Gao Huafeng Wang Zhou Feng Mingxia Fu Chennan Ma Haixia Pan Binshen Xu Ning Li

作者单位: School of Software,Beihang University of Beijing,Beijing 10083,China School of Software,Beihang University of Beijing,Beijing 10083,China;North China University of Techn Research Institute of Petroleum Exploration and Development of PetroChina,Beijing 100083,China

会议类型: 国际会议

会议名称: 第七届全国模式识别学术会议(The 7th Chinese Conference on Pattern Recognition,CCPR2016)

会议地点: 成都

会议语种:英文

页码: 161-172

在线出版日期: 2016-11-03（万方平台首次上网日期，不代表论文的发表时间）