摘要
油藏地质剖面图是了解油藏变化起伏形态、地层分布特征、油水关系及物性特征的重要图件,能为系统分析区域或局部的油藏地质条件、正确指导油藏资源开发利用和优化管理提供依据。随着计算机技术的发展,实现对已有油藏地质剖面图的矢量化对油藏地质剖面的深度利用具有重要意义。针对油藏地质剖面图的特点,对Canny边缘检测算法进行了改进,引入了k-means聚类分割,有效地删除了图中的各种图例信息,较好地保留了油藏地质剖面的水平层位轮廓,解决了油藏地质剖面图岩性符号多而难以获得完整的水平层位轮廓、矢量化效果差等问题。实验结果表明,该方法对于各种复杂油藏地质剖面图的矢量化效果好、效率高,并在随钻地质综合解释中得到了很好的实际应用。
The image of the oil reservoir geological profile is an important picture to understand the oil reservoir fluctuation,characteristics of stratigraphic distribution,oil-water relationship and characteristics of physical properties,which can provide a basis for systematically analyzing the regional or local oil reservoir geological conditions and properly guiding the development and utilization of oil reservoir resources and optimizing management.With the development of computer technology,it is very important to realize the vectorization of the geological profile of oil reservoirs.According to the characteristics of oil reservoir geological profile,we improve the Canny edge detection algorithm,introduce k-means clustering segmentation,and effectively remove the various legend information in the figure,which can keep the contours of the horizontal horizons of the oil reservoir geological profile,and solve the problem that the lithologic symbols of the oil reservoir geological profile are large and difficult to obtain the complete horizontal horizon profile and the vectorization effect is poor.The experiment shows that the proposed method is effective and efficient in the geological profile of various complex oil reservoirs,and it has been applied in the comprehensive interpretation of drilling and geology.
作者
段友祥
李宁宁
孙歧峰
李钰
DUAN You-xiang;LI Ning-ning;SUN Qi-feng;LI Yu(School of Computer and Communication Engineering,China University of Petroleum,Qingdao 266580,Chin)
出处
《计算机技术与发展》
2018年第8期170-174,共5页
Computer Technology and Development
基金
国家十三五"重大专项"(2017ZX05009-001)
关键词
地质剖面图
边缘检测
地层
矢量化
聚类
geological profile
edge detection
stratigraphy
vectorization
clustering