【作者】 吕林；

【导师】 陈宝心；

【作者基本信息】 武汉理工大学 ， 采矿工程， 2011， 硕士

【摘要】 岩体中含有大量性状不同、形态各异的节理裂隙,它们不仅控制着岩体力学性质,而且控制岩体的破坏形式,通常认为露天台阶岩体的爆破破碎实际上是在岩体已被各种地质构造弱面和前次爆破弱面切割成具有某种分布规律的天然岩块基础上的再次破碎过程。天然岩块的块度的测算方法一般有测线法、取岩芯法以及由爆破结果反求天然块度方法等,这些方法虽能较好的对天然岩块块度进行定量评价,但都存在现场实测工作量大、实施困难等缺陷。基于摄影图像处理技术的天然岩块块度分析方法则可以很好的弥补上述方法的诸多缺陷。文章在总结岩体天然块度计算的理论和方法的基础上,深入研究了当前数字图像处理主要技术及其应用,成功的开发了岩体节理裂隙图像的分析处理软件。该软件在MATLAB环境下采用面向对象方法开发,实现了岩体图像的分辨率推导、岩体图像的滤波平滑除噪预处理、岩体图像的边缘检测、图像的多尺度梯度信息的提取、岩块分割、岩块几何尺寸的量算(面积)、块度曲线的绘制以及块度信息的存档等功能。在进行岩体图像识别和分析处理的过程中,图像分割是最关键的环节之一。针对岩体图像的特点,对岩块的自动分割进行着重研究,在传统的分水岭变换的基础上,结合多尺度梯度算子和形态学重建等技术,建立了一种新的改进的分水岭分割算法模型,很好的解决了岩体图像的分割。最后,利用现场采集的台阶坡面图像对系统进行性能测试,结果证明系统具有较高的测试精度,分析结果可靠准确,分析过程快速高效。更多还原

【Abstract】 There are a lot of joints and cracks which have different shape and nature in the rock mass. They not only control the mechanical properties of rock mass, and control the destruction form of rock mass. Generally consider the bench rock blasting is just a second breaking to the rock which had been cutted into pieces with a certain distribution by various geological structure weak surfaces and previous blasting. The methods of the natural rock fragmentation measuring are "scanline method","core analysis method"and "inverse calculation method from blasting result" and etc. Although both of these methods can quantitative evaluate the fragment of natural rock accurately, there are many defects such as too heavy in-situ test work, too many implementation difficulties and etc. The analysis method of the natural rock size based on the image processing technology can mend these defects.This paper summarizes the theory and method of the natural rock fragmentation, deeply researchs the digital image processing technology and its application and develops a image analysis system for the roak mass image successfully. This system is developed by using object-oriented method in the MATLAB environment.Which realizes the resolution derivation of the rock mass image, pretreatment of the rock mass image, edge detection, multi-scale gradient information extraction, image segmentation, the rock geometry size calculation (area), size distribution curve drawing, storing of the size information and etc.Image segmentation is one of the mostimportant steps during rock mass image recognition and analysis processing. The algorithm based on watershed transform is improved. According to the characteristics of rock image,the watershed transform, multi-scale gradient operator and morphological reconstruction are integrated to segment images commendably.Finally, the system performance is tested by the bench rock mass image from site. Results proves that the system has high accuracy. The analysis results are reliable and the analysis process quickly and efficiently.更多还原