三维岩心图像裂缝自动识别

摘要

In the field of oil geology, crack recognition is considered as a very important part when relating to oil reserves and heterogeneous reservoir study. Using three-dimensional reconstruction technology, this essay overlapping core CT image series, and gain the target's fitting plane by least square method. By calculating the ratio of minimal circumscribed sphere radius and equivalent sphere radius, shape fac tor and the ratio of the longest side and shortest side of approximate minimum circumscribed cuboid, it enables the automatic recognition of core crack. The experimental results show that the related method improves the recognition rate of core crack recognition.%在石油地质领域,裂缝识别是油藏和储层非均质研究的重要内容,对油气开发和原油采收率的提高具有重要意义.文章对岩心CT序列图像进行三维重建,采用最小二乘法得到三维目标的拟合平面,通过计算三维目标最小外接球半径与等效球半径的比值,形状因子以及近似最小外接长方体最长边与最短边的比值,实现岩心裂缝的自动识别.实验结果证实,该方法能够较好地识别出三维图像中的裂缝.