成像测井中的裂缝自动识别方法

Automatically Extract Fracture Parameters from Resistivity Images by Using Hough Transform

现有的成像测井解释软件大多是以关键点拟合来提取裂缝的。研究了利用霍夫 (Hough)变换自动识别裂缝的方法及技术。由于裂缝在成像测井图中表现为单周期的正弦曲线 ,通过对含有已知参数的正弦曲线图像进行识别 ,证明该方法具有准确识别正弦曲线及其参数的能力 ,即使在原图像信噪比很低的情况下仍能给出理想的识别效果。实际资料处理表明 。

Most of present imaging interpretation softwares only detect the fractures mainly by fitting the key points selected by interpreters. This paper gives an automatic method for picking up the fractures without interference. Due to the characters of fracture's single cycle sinusoidal curves, it has been proved that Hough transform could accurately detect the fractures and give their parameters by identifying the parameter known images. And also a satisfying effect could still be gained even though the signal noise ratio in the original images is lower. A case study shows that this approach could always give a correct result no matter there are single or cross fractures in the borehole walls logged by resistivity imaging logging.