基于最大类间方差遗传算法的岩心CT扫描图像分割方法

ROCK CT SCAN IMAGE SEGMENTATION BASED ON THE OTSU GENETIC ALGORITHMS

目前，CT技术在油气勘探领域中应用越来越广泛，尤其是在岩心分析方面。本文在分析岩心CT扫描图像特征的基础上，提出一种消除岩心CT图像背景影响的分割方法。首先对岩心CT扫描图像统计灰度直方图，根据该直方图确定一个灰度值，并用该灰度值代替图片的背景灰度值，然后利用最大类间方差法结合遗传算法确定最优闲值并对原始CT扫描图像处理。结果表明，该方法能消除岩心CT扫描图像背景的影响，将目标从CT扫描图像中提取出来，为岩心分析提供很好的基础。

At present,The CT scan has been used in oil exploration more and more widely,especially in rock data analysis. Based on study of rock CT scan image characters,a method of rock CT scan image segmentation which can eliminate the effect of the image background is proposed in this article.First,grayscale value histogram of the rock CT scan image is made to give a grayscale value.And grayscale values of the image back ground are replaced by this grayscale value.Then the optimal threshold is qualified using the genetic algorithms on the base of otsu,and the rock CT scan image is processed.The result shows that by using this method the effect of rock CT scan image background can be well eliminated,which will be used for rock data analysis.