铸体薄片图像分析法求取储层孔隙度  

Strike the Reservoir Porosity Using the Casting Sheet Image

作　　者：刘庆利[1] 吴国平[1] 胡剑策[1] 侯卫国 

机构地区：[1]中国地质大学信息工程学院,湖北武汉430074 [2]中石化西北分公司,新疆乌鲁木齐830011

出　　处：《测绘科学技术学报》2009年第1期69-71,75,共4页Journal of Geomatics Science and Technology

基　　金：国家自然科学基金资助(40674069)

摘　　要：在石油勘探工程中,储层孔隙度是进行油气预测、油气储量计算的重要参数。根据测井铸体薄片图像资料,在传统BP网络模型的基础上提出了一种基于灰色关联约束的BP神经网络改进模型。对铸体薄片图像进行孔隙识别、统计,进而自动判读储层孔隙度。研究和实验表明,该方法对储层孔隙度的分析、判读更具合理性和实用性。In the area of the oil exploration engineering, the reservoir porosity is important parameters to forecast and calculate the oil and gas. In terms of the casting sheet images, this paper presented a BP neural networks pattern based on grey relationship, then used it to analysis and identify the pore and to calculate the reservoir porosity auto- matically. The research and experiments showed that the method was more reasonable and practical to analysis and calculate the reservoir porosity.

关 键 词：铸体薄片 图像识别 灰色关联约束-BP神经网络 孔隙度判读 

分 类 号：P258[天文地球—测绘科学与技术]