基于数字图像处理的岩体结构面迹线检测

Rock mass discontinuity trace detection based on digital image processing

岩体结构面信息的准确获取和识别对于研究岩体的力学性质具有重要作用。基于数字图像处理技术,利用MATLAB软件编程对图像进行灰度化处理、图像滤波、感兴趣区域选取、边缘检测等操作,得到边缘检测图像,再通过线分段连接的方法形成结构面迹线。利用该方法对云南自走铁矿露天边坡露头结构面进行了检测,并将其与概率霍夫变换检测、人工绘制等方法进行对比,结果表明,新方法检测得到的结构面迹线的准确率为91%、误检率约为3%、漏检率约为9%,得到的迹线较为完整、连续且位置准确,与人工绘制的差距很小,从而为矿山调查结构面产状提供了一种较为便捷准确的方法。

The acquisition and identification of the accurate information about structural plane of rock mass play an important role in studying mechanical properties of rock mass.Based on digital image processing technology,Matlab software was employed to carry out image graying,image filtering,interested region selection,edge detection and other operations to obtain the edge detected image,and then form the structural plane trace using line segmentation connection method.This method was applied to detect the structural plane of the outcrop slope in an open-pit iron mine in Yunnan province.By comparison of the results with that from probability Hough transform detection and manual drawing,it was found that the accuracy rate of structural plane trace detected by the new method was 91%with the false drop rate of about 3%and the omission rate of about 9%.A complete,continuous trace with accurate location was produced by the novel method and there was a minimal discrepancy to the manual drawn.Consequently,it provides a convenient and fast and accurate approach for mines to investigate the occurrence of structural plane.