摘要
有限应变测量是研究岩石变形特征的重要研究内容。为了提高岩石有限应变测量的精度与速度,改变手工测量岩石有限应变耗时、枯燥的状况,应用多重图像限制的自动种子区域生长算法,以大青山地区为例,对如变形区内的石英砂岩有限应变进行了自动测量。该算法综合多个图像、利用色彩信息,不需要特殊设备即可对颗粒进行自动识别。算法还在叠加后的图像中,应用了修改的Canny边缘检测算法增强颗粒边界与内部的对比度。识别结果与手工测量的颗粒,在质心坐标、面积、长轴、短轴、有限应变测量方面具有很大的相似性。
Finite strain determination is an important research for deformation characteristics of rocks. In order to improve the precision and the velocity, and ameliorate the status which is laborious and time consuming methods availability for manual finite strain determination, we take automatic method of finite strain determination by MICASRG algorithm, which is short for Multiply Images Constrained Automated Seeded Region Growing, and set the quartz sandstone of Daqing Mountain as an example. Based on multiply images and information of color, the MICASRG can recognize the grain without the necessity for the special equipment. In addition, this method enhances the contrast gradient between the inner and the boundary of the grain by the Canny operator. The results of automatic identified and handcrafted have the great comparability on centroid, area, major axis, minor axis and finite strain determination, and reflect the feasibility of this method. We also give a satisfied explanation about the poor correlation on orientation of the major axis between auto-identified and handcrafted.
出处
《地质力学学报》
CSCD
2014年第2期159-164,共6页
Journal of Geomechanics
基金
国家科技重大专项项目(2011ZX05023-001)
关键词
构造变形
应变测量
多图像
石英砂岩
自动化方法
structural deformation
finite strain determination
multiply images
quartz sandstone
auto-method