基于不完全路径形态学算法的电成像测井资料缝洞识别与参数评价研究  

Identification and Quantitative Evaluation of Fracture-vuggy from Electric Conductivity Images Based on Incomplete Path Morphologic Algorithm

作　　者：丁娱娇 金凤鸣 沈金松[2] 孙帅[2] 啜晓亚 李新耀 Ding Yujiao;Jin Fengming;Shen Jinsong;Sun Shuai;Chuo Xiaoya;Li Xinyao(Institute of Oil Exploration and Development,Dagang Oilfield,CNPC,Tianjin 300280,China;College of Geophysics,China Petroleum University(Beijing),Beijing 100100,China)

机构地区：[1]中国石油大港油田勘探开发研究院地球物理学院,天津300280 [2]中国石油大学(北京)地球物理学院,北京100100

出　　处：《工程地球物理学报》2025年第2期247-259,共13页Chinese Journal of Engineering Geophysics

基　　金：国家自然科学基金面上项目(编号:42074127);中石油重大专项(编号:2023ZZ15YJ02)。

摘　　要：基于电成像测井的电导率图像数据,用不完全路径形态学算法从成像测井电导率图像提取不同尺度和不同形态的裂缝和溶蚀孔洞,结合密度聚类算法与改进的Hough变换和正弦函数拟合,实现了裂缝和溶蚀孔洞的识别和缝洞孔隙度的定量评价。文中首先用图像处理中的快速行进(Fast Marching Method,FMM)对电成像测井电导率图像的空白带进行了插值、畸变区域修复与噪声压制,使之更完整地反映井壁裂缝与溶蚀孔洞的分布;其次,引入图像处理中的不完全路径形态学算法从电导率图像识别与标注和缝洞对应的电导率异常;再次,分析了如何优化密度聚类融合算法参数,将断续分布及不完整的缝洞组合成正确的缝洞分布,举例说明了不同参数的应用效果;还有,基于改进的Hough变换和正弦函数拟合方法对识别的裂缝与溶蚀孔洞进行拟合,提取裂缝参数,用图像矩和椭圆函数拟合提取溶蚀孔洞参数,并画出裂缝的倾向与倾角分布图。最后,采用数值模拟和实际测量的电导率成像数据验证了文中算法对自动提取缝洞参数的有效性。Based on electric conductivity images,the incomplete path morphology algorithm is used to extract fractures and dissolved pores with different scales and morphologies from the electric conductivity images in imaging logging.Density clustering algorithm,improved Hough transform and sinusoidal function fitting is combined to quantitatively evaluate fractures and dissolved pores.Firstly,the blank bands of the conductivity image of the electric imaging logging are interpolated,distorted areas are repaired and noise suppressed by FMM,so as to reflect the distribution of the fractures and dissolved pores completely.Secondly,the incomplete path morphology algorithm is used to identify and label the conductivity anomalies from the conductivity image These anomalies correspond to cracks and dissolved pores respectively.Thirdly,the parameters of the density clustering fusion algorithm are analyzed to combine intermittent fractures into complete fractures,and examples are given to illustrate the effects of applying different parameters.Fifthly,fractures are fitted based on the improved Hough transform and sinusoidal function fitting method,and dissolved pores parameters are extracted based on image moments and elliptic function fitting.Finally,the effectiveness of the algorithm for automatically extracting the fracture-vuggy parameters is verified using numerical simulations and actual measured conductivity imaging data.

关 键 词：不完全路径形态算法 电成像测井电导率图像 裂缝与溶蚀孔洞 HOUGH变换 密度聚类 

分 类 号：P631.8[天文地球—地质矿产勘探]