数字钻孔全景影像中结构面特征智能识别方法

数字钻孔全景影像是识别深部岩体结构面的主要方法，传统解译方法主要依靠手动操作完成，其结果存在较强的人为性、主观性和较大误差。针对数字钻孔全景影像技术，提出了一种考虑结构面图像灰度分布特点的智能识别新方法。通过钻孔图像的灰度化、降噪预处理后利用结构面定位信号特征值D定位结构面，获取其所在区域图像；再进行边缘检测、阈值分割及形态学等处理，通过拟合边缘曲线并结合数学表征方法实现了结构面特征(产状、隙宽等几何参数)的智能化识别。通过对地下水封洞库工程深部岩体数字钻孔影像的实例分析，运用本智能识别方法获得结构面特征信息相对于传统人工识别方法，其结果的准确性和客观性更强、批量识别效率更高。这对这数字钻孔影像的智能、快速识别具有一定的参考价值。 

Intelligent identification method for rock discontinuities properties by digital borehole panoramic images

Digital borehole panoramic imaging technique is the main method to identify the discontinuities properties of deep rock mass. The traditional identification method mainly relies on manual operation, therefore the results are artificial, subjective and inaccurate. A new intelligent identification method is proposed considering gray level distribution of rock discontinuities for the digital borehole panoramic imaging technique. The panoramic images are firstly pre-processed by graying and noise reduction, and positioned the joints by the eigenvalue, D, to obtain the joints region images. The joints images are processed after edge detection, threshold partition and morphology processing. Finally, intelligent identification of discontinuities characteristics are realized by joints edge curve fitting and mathematical calculation, which are dip direction, dip and aperture. Compared with a traditional artificial method, the proposed technique is more accurate and objective, and more efficient in batch images identification.