论文部分内容阅读
地震数据的精确解释和分析非常依赖于所使用算法的稳定性。我们着重于盐丘地震勘探的稳定检测。我们讨论一个基于地震成像中最佳结构特征属性排序分类的盐丘探测新模型。该算法克服了现有的基于结构属性技术的局限性,因为该技术非常依赖于盐丘固有的地质属性与盐丘检测所用的属性数量。该算法综合了灰度共生矩阵(GLCM)属性,Gabor滤波器,以及含有使用属性特征排序信息协方差矩阵的本征结构等属性。将排序前列的属性组合起来形成一组最优的特征集,以保证算法即使在沿盐丘边界没有强反射层的情况下也能有效。与现有的盐丘检测技术相比,本文的算法稳定和计算高效,并能处理小尺度特征集。我用荷兰F3地块评价该算法的性能。实验结果表明,本文提出的基于信息理论的工作流程用于检测盐丘,其精度优于现有盐丘检测技术。
The precise interpretation and analysis of seismic data is very dependent on the stability of the algorithm used. We focus on the stable detection of salt dome seismic exploration. We discuss a new salt dome detection model based on the sorting of the best structural features in seismic images. This algorithm overcomes the limitations of the existing structure-based techniques because the technique relies heavily on the inherent geological properties of the salt dome and the number of attributes used in the salt dome detection. The algorithm combines attributes such as GLCM, Gabor filters, and intrinsic structures that contain covariance matrices that use attribute signature ordering information. Combine the top-ranked attributes to form a set of optimal feature sets to ensure that the algorithm is valid even without strong reflections along salt dome boundaries. Compared with the existing salt dome detection technology, the proposed algorithm is stable and computationally efficient, and can process small scale feature sets. I use the Netherlands F3 plots to evaluate the performance of the algorithm. The experimental results show that the proposed workflow based on information theory is used to detect salt domes and the accuracy is better than the existing salt dome detection techniques.