基于空间标记的岩心孔隙并行分割  

Parallel Segmentation of Core Pores Based on Spatial Markers

作　　者：陈国军[1] 李胜 CHEN Guo-jun;LI Sheng(School of Computer Science and Technology,China University of Petroleum(East China),Qingdao 266580,China)

机构地区：[1]中国石油大学(华东)计算机科学与技术学院,山东青岛266580

出　　处：《计算机技术与发展》2020年第12期142-147,共6页Computer Technology and Development

基　　金：国家油气重大专项(2016ZX0502)。

摘　　要：由于传统图像分割方法只针对岩心图像二维信息进行处理,忽略了岩心图像在三维空间中的联系,导致岩心图像孔隙骨架分割的准确性不高。岩心内部空间分布复杂,图像标注工作量大、效率低,不适用深度学习的方法进行分割。提出了一种基于空间标记的岩心孔隙骨架并行分割方法,采用最大类间方差法计算岩心图像的预分割阈值。应用多线程技术对图像并行阈值分割和连通域标记,并结合二路归并思想将多帧图像空间相邻的连通域合并为多个独立的连通体。根据边界限定准则剔除孤立的连通体,并以实验孔隙度为指导迭代优化分割阈值。实验结果表明,采用并行标记的方法耗时为普通方法的三分之一,利用图像间的三维联系实现了岩心图像三维分割,生成的数字岩心三维模型有效去除了孤立的孔隙像素,提高了模型的准确性。The traditional image segmentation method only deals with the two-dimensional information of the core image and ignores the relationship of the core image in the three-dimensional space.As a result,the accuracy of the core skeleton pore skeleton segmentation is not high.The internal space distribution of the core is complicated,the image labeling workload is large,and the efficiency is low,so it is not suitable for deep learning to perform segmentation.A parallel segmentation method of core pore skeleton based on spatial markers is proposed.The maximum segment variance method is used to calculate the pre-segmentation threshold of core images.Multi-threading technology is applied to image parallel threshold segmentation and connected domain labeling,and the two-way merge idea is used to merge adjacent connected domains in multiple image space into multiple independent connected bodies.The isolated connected body is eliminated according to the boundary definition criterion,and the segmentation threshold is iteratively optimized based on the experimental porosity.The experiment shows that the parallel labeling method takes one-third the time of the ordinary method.The three-dimensional segmentation of the core image is achieved by using the three-dimensional relationship between the images.The generated three-dimensional digital core model effectively removes the isolated pore pixels and improves the model accuracy.

关 键 词：数字岩心 连通域 空间标记 多线程 孔隙度 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]