基于模糊算法的岩心图像层理特征提取

在岩心图像的分析中,提取图像的层理特征,对于识别岩石的层理类型,完成对获取的岩心图像进行自动识别十分重要。本文介绍了Pal.King算法,并且针对岩心图像的多数像素点处在低灰度级的特点,使用最大类间方差法(Otsu)得到Tr变换中的最佳隶属度的阈值,根据阈值建立自己的隶属函数G和G~(-1);同时对Tr变换增强后的空间域图像进行平滑处理,减少噪声对层理边缘提取的影响。实验结果表明,该算法提取岩心图像层理特征效果较好。

Feature extraction of core image based on fuzzy algorithm

To extract feature of image in core image analysis is important to identify the texture type and access the core image automatic identification.The paper introduces the Pal.King algorithm.Since majority of pixels in core image are in gray level,the paper makes use of Otsu to get optimal membership threshold in Tr transform,and constitutes its own membership function of G and G~(-1) according to optimal threshold.Simultaneously,we need to smooth the spatial domain image which has been Tr transformed to reduce the impact of noise on the texture edge extraction.The result shows that the algorithm is effective in extracting core image texture feature.