基于改进的区域生长和Hough变换的钻孔图像水平裂隙识别

利用前视钻孔摄像系统(Axial View Borehole Camera System, AVBCS)获取地质钻孔孔壁图像并分析地质结构是一种重要的地质勘查方式。针对目前钻孔孔壁图像基本由人工判读的现状，提出了一种基于改进的区域生长和Hough变换的钻孔孔壁图像水平裂隙识别算法。首先，利用三种图像预处理算法提高钻孔图像中裂隙区域的对比度；然后，结合OTSU算法和圆形度自动提取种子点，并利用双阈值准则进行中心区域生长，以消除中心干扰区域；最后，利用Hough变换算法完成裂隙的检测与识别。实验结果表明本文提出的裂隙识别算法对于不同的水平裂隙和离层图像都可以准确、自动的检测出裂隙。

Horizontal Fracture Recognition of Borehole Image Based on Improved Region Growing and Hough Transform

Obtaining the borehole wall image and analyzing geological structure based on the Axial View Borehole Camera System (AVBCS) is an important way of geological exploration. In order to solve the present situation that the borehole wall images are mainly manual interpretation, horizontal fracture recognition algorithm based on improved region growing and Hough transform is proposed. Firstly, the contrast of fracture region in the borehole image is improved by three image preprocessing algorithms. Then, the automatic extraction of seed points is realized based on OTSU and circularity, and the central interference region is eliminated by region growing of double threshold criterion. Finally, use Hough transform to detect and recognize the fracture. The experimental results show that the proposed fracture recognition algorithm can detect different horizontal fractures and abscission layers accurately and automatically.