【摘要】 在地震资料解释工作中,一般的地震相分析工作效率较低,有时难度较大。通过计算灰度共生矩阵GL-CM,从数学上描述小范围数据区内像素值的分布,量化地震反射的空间组织结构。将构造属性与神经网络分类技术相结合,通过训练网络和质量控制,利用波形相似性和地震属性进行地震相分析,可用于对三维地震资料进行地震相的划分,得到三维地震相分类数据体。这种方法减少了许多耗时的工作,使解释人员能够集中精力研究地震相,并将其综合成地质成果图。更多还原

【Abstract】 Conventionally seismic facies analysis is implemented manually which is time-consuming, and in some cases, is difficult to map different seismic facies consistently. Calculation of grey-level co-occurrence matrices (GLCM) helps mathematically describe the distribution of pixel values within a subregion of data, and effectively quantify the spatial structures of seismic reflections. Thus seismic textural analysis and neural network can be successfully combined. After training and QC, seismic facies are classified in terms of waveform similarity and seismic attributes. 3-D seismic facies classification data volume can be generated.更多还原