摘要
岩石的孔隙、喉道等特征是地质人员进行判断储层特征的重要参数。因岩石图像具有较高的复杂度,图像纹理相似,因而对岩石图像进行分析时有一定的难度。人工进行分析时,容易因为各种原因出现误差。本文提出利用模糊C均值算法(FCM)对鄂尔多斯盆地岩石铸体薄片进行聚类分析,实验结果能很好的将岩石铸体薄片中孔隙与岩石背景区分出来,为后期的岩石自动识别与分类奠定了基础。FCM是基于划分的一种非监督聚类算法。
The characteristics of rock pores and throats are important parameters for geological personnel to judge reservoir characteristics. Due to the high complexity of the rock image and the similar texture of the image,it is difficult to analyze the rock image. When manual analysis is performed,errors may easily occur due to various reasons. This paper proposes the use of fuzzy C-means algorithm（ FCM） to cluster analysis of the rock casting flakes in the Ordos Basin. The experimental results can well distinguish the pores and rock backgrounds of the rock casting flakes and lay a foundation for the later automatic identification and classification of rocks. FCM is an unsupervised clustering algorithm based on partitioning.
作者
程国建
宋博敬
CHENG Guojian;SONG Bojing(School of Computer Science,Xi＇an Shiyou University,Xi＇an 710065,China)
出处
《智能计算机与应用》
2018年第4期78-80,84,共4页
Intelligent Computer and Applications
基金
陕西省工业科技攻关项目(2015GY104)
关键词
岩石铸体薄片
非监督聚类算法
模糊聚类
rock casting flake
unsupervised clustering algorithm
fuzzy clustering
