成像测井井壁图像裂缝自动识别

利用数学形态学中的腐蚀和膨胀技术以及图识别的方法实现了对简单的地层裂缝信息的自动识别，数学形态学是以随机集论，积分几何，拓扑学，近代代数为基础的用于数字图像处理和识别的学科，大部分形态学运算都定义在腐蚀和膨胀的基本运算的基础上，通过运算处理达到改善图像质量，描述和定义图像的各种几何参数和和特征如面积，周长，连通性，颗粒度，骨架和方向性等目的，CNPC测井项目组利用上述方法开发的成像测井解释系统软件支持斯伦贝谢，阿特拉斯，哈里伯顿以及多种国像测井仪器，具有一定的抗噪能力，应用效果良好。

A Math-morphological Approach for Automatic Fracture Recognition on Sidewall Images

Erosion algorithm and dilation algorithm in math morphology together with image processing are successfully used for automatic simple fracture recognition on side wall images. Math morphology, based on stochastic set theory, integral geometry, topology and modern algebra, is a subject for digital image processing and recognition. Most morphological approaches are on the basis of erosion and dilation algorithms, with which we can improve image quality, describe and define various geometry parameters and features of images, such as area, circumference, continuity, grain size, matrix and directivity, etc.. With the method mentioned above, imaging log interpretation software developed by CNPC logging project team helps to realize the erosion and dilation algorithms, and also processes log data from FMS2, FM1, STAR, EM1, CBIL as well as some other China made imaging logging tools. Applications prove this approach is able to provide more reliable calculations, in other words, more precise fracture images.