论文部分内容阅读
在火成岩岩性识别过程中,常规测井资料主要反映不同岩性的成分特性,对岩石结构的信息反映较弱,而成像测井资料能够直观的反映岩石结构、构造等特征。综合二者的优点,本文基于计算机图形学算法[1]与支持向量机(SVM)分类器原理,提出了一种利用成像测井图像纹理特征自动识别火成岩岩性的新方法。实践表明,该方法与常规测井方法相比,识别率得到明显提高,且识别结果与实际情况相符合。
During the lithologic identification of igneous rocks, the conventional logging data mainly reflect the compositional characteristics of different lithologies and reflect weakly the information of rock structures. Imaging logging data can intuitively reflect the characteristics of rock structure and structure. Combining the advantages of the two, this paper presents a new method to automatically identify the igneous rock lithology by using image texture feature of imaging log, based on the principle of computer graphics [1] and support vector machine (SVM) classifier. Practice shows that compared with the conventional logging method, the recognition rate of the method is obviously improved, and the recognition result is in accordance with the actual situation.