摘要
数字露头的研究方法主要有激光雷达技术和数字摄影测量技术,构建的实景三维模型,可进行剖面量测、岩层岩性分析、岩层走向分析、岩层叠覆情况分析、构造和沉积信息等多种地质分析,可为重现的虚拟实测(厚度、层/面理和线理等)及定量统计奠定可靠的数据基础,可为碳酸盐岩岩溶缝洞型储层的预测和油藏开发提供重要的研究思路和依据。通过查阅相关文献,调研了目前已开展的碳酸盐岩剖面数字研究实例:四川盆地川东–鄂西三叠系飞仙关组和鄂西利川见天坝剖面长兴组,塔里木盆地塔中地区良里塔格组和柯坪坝地区肖尔布拉克组等数字露头剖面。开展的研究内容包括:基于多尺度区域卷积神经网络的露头孔洞自动提取;基于数字露头表征的岩溶缝洞组构特征和碳酸盐岩储集层三维地质建模;基于无人机倾斜摄影的碳酸盐岩生物礁露头多尺度非均质性表征以及三维地质建模。
The research methods of digital outcrop mainly include lidar technology and digital photogrammetry technology. The 3D model constructed by the real scene can be used for a variety of geological analysis, such as section measurement, rock lithology analysis, rock strike analysis, rock overlying condition analysis, structure and sedimentary information. It can lay a reliable data foundation for repeated virtual measurement (thickness, bedding/plane bedding and linings, etc.) and quantitative statistics, and provide important research ideas and basis for the prediction and reservoir development of carbonate karst fracture-vuggy reservoirs. By referring to relevant literature, the digital research examples of carbonate profiles have been investigated, including the Feixianguan Formation of Triassic in eastern Sichuan and western Hubei, the Changxing Formation of Jianchangba section in Lichuan, western Hubei, the digital outgrowth profiles of Lianglitage Formation in Tazhong area, Tarim Basin, and Xiao Erbulak Formation in Kepingba area. The research contents include automatic extraction of outcrop holes based on multi-scale regional convolutional neural network;The fabric characteristics of karst fractures and caves based on digital outcrop characterization and 3D geological modeling of carbonate reservoirs;Multi-scale heterogeneity characterization and 3D geological modeling of carbonate reef outcrop based on UAV tilt photography.
出处
《地球科学前沿（汉斯）》
2024年第5期647-661,共15页
Advances in Geosciences
