Digital rock segmentation from micro-CT/SEM data by using convolutional neural network

Pore-scale digital images are usually obtained from micro computed tomography (micro-CT) data that has been segmented into void and grain space. Image segmentation is a critical process for digital rock analyses that can influence pore-scale characterisation studies and/or the numerical simulation of petrophysical properties. This project is to study on the sandstone micro-CT image segmentation by using convolutional neural network.

孔隙尺度数字图像通常源自经分割为孔隙空间与颗粒空间的显微计算机断层扫描（micro computed tomography，micro-CT）数据。图像分割是数字岩石分析的关键环节，其会对孔隙尺度表征研究及岩石物理属性的数值模拟工作产生影响。本项目旨在采用卷积神经网络开展砂岩显微计算机断层扫描图像分割相关研究。