钻孔成像中结构面自动判读技术研究

Study of automatically extracting structural plane parameters from borehole images

通过钻孔数字成像系统得到地质钻孔的孔壁数字图像并研判地下地质结构是一种重要的地质勘查手段。对钻孔成像中的结构面信息的判读通常是由人工完成的,不同的人判读结果可能存在差异。钻孔成像自动辨识结构面系统首先对图像进行RGB(red红,green绿,blue蓝)到HSV(hue色相,saturation饱和度,value明度)颜色空间的转换,使得结构面信息突出显示出来,以利于图像的结构面提取;再对转换后的S分量进行图像滤波及增强等预处理,然后对图像进行分割,对分割后的图像进行数学形态学中的二值图像边缘提取和细化处理,快速分离出结构面信息,对结构面边缘进行有效的提取;最后采用正弦曲线的Hough变换对图像边缘点进行识别,可得到钻孔结构面的产状和厚度参数。以实际钻孔数字图像为例,经过钻孔成像自动辨识结构面系统提取了其中的4个结构面的空间参数,与人工判读结果比较,自动辨识得到的参数更精确,结果具有唯一性。

Collecting the digital images of geological drilling holes and interpreting the underground geological structure is a new method of geological prospecting.The interpretation of structural planes on drilling holes images is usually done by hand;and results of interpretation may be quite different for different people.Automatic identification system of structural plane（AISSP） can process the digital images of drilling holes to get structural plane parameters automatically.First of all,the images conversion from RGB color space to HSV color space should be done to highlight the structural plane region.Secondly,the S component images are filtered and enhanced for the next processing step.After the image segmentation,the binary image edge will be extracted and thinned by image morphology method;and edges of the structural plane are effectively extracted.Finally,structure plane altitude and thickness parameters are obtained by the Hough transform of sine curve and identified edge points.As example,a real digital drilling hole picture is analyzed by AISSP and the space parameters of the 4 structural planes appearing in the picture are calculated.Compared with the results interpreted by people,the results from AISSP are more precise and unique.