井壁成像测井资料目标体自动检测方法研究

根据现有的多种图像自动分割算法(如直方图统计法、判断分析法、P 参数法、微分直方图法、灰度差直方图法、熵参数法、基于过渡区法和双阈值法)及测井目标体的特点,给出了一种适合于井壁成像测井资料目标体自动检测的方法,实现了缝洞目标体的定量检测。运用这种方法对井壁声测井图像进行了实例处理分析,并与其他方法的处理结果进行了对比,结果表明该方法可行、有效,具有实用性。

Automatic detection of borehole image logging target

The existing image segmentation algorithms includes histogram statistic dividing, variance analysis dividing, P-tile dividing, difference histogram dividing, difference image histogram dividing, entopic dividing and the transition region dividing. On the basis of these algorithms, a new automatic segmentation algorithm for borehole image logging was presented in consideration of the formation characteristics. The algorithm makes fracture and hole measurable. The calculated exemplum indicates that this new algorithm is more available, practicable and effective than any other algorithm.