中文摘要：

基于 CT图像的材料孔隙率分析是评价岩土多孔介质的重要手段。当前孔隙率的计算主要针对 CT单张图像采用阈值分割法计算，然而，由于阈值分割法在确定孔洞与固体基质灰度门槛值时的不确定性，即便针对相同的 CT图像由于不同的阈值分割方法而使计算结果相差较大。为了克服阈值分割法确定材料孔隙率的缺陷，探索一种基于 CT图像确定岩土介质孔隙的新方法。首先，借助于数字地面高程基本思路，提出基于 CT图像灰度水平的孔隙率计算方法，从图像的 CT数出发将 CT数分成两类，一类属于孔洞，另一类属于基质，对孔隙计算相关公式进行推导。其次，将灰度水平法应用于土石混合体材料的 CT单张图像孔隙率计算当中，所得计算结果并与两种阈值分割法进行对比研究，证明了灰度水平方法的优势与可行性。最后，在得到试样孔隙率的基础上，从试样细观孔隙演化角度对试样的变形破坏机理进行了分析。

英文摘要：

Porosity analysis based on CT images is a significant method to evaluate the geotechnical po-rous media,and the porosity is currently calculated by threshold segmentation method for single CT image. However, because of the uncertainty of segmentation method in determining the gray threshold value for pores and matrix, even for the same CT image the calculation results from different segmentation are quite different. To overcome the drawback of gray threshold methods, a new method, called here the gray level method was proposed firstly. Using this method, CT number of images can be divided into two camps, one belongs to the pores, and the other belongs to matrix. Then, the proposed method is used to the po-rosity calculation of rock-soil aggregate. The estimated porosity result was compared with two segmentation methods. A close agreement with the conventional method was found and reliability of the proposed method has been proved. At last,based on the estimated porosity using the proposed method,the deformation and failure process of rock-soil aggregate is analyzed from aspect of meso-porosity evolution.