基于图像分割+SVM的微观剩余油类型自动识别

Automatic Recognition of Micro Remaining Oil Type Based on Image Segmentation+SVM

剩余油主要吸附在地下岩石中,研究剩余油的载体是解决剩余油问题的关键,而油层孔隙中剩余油的微观赋存状态是指导剩余油挖潜的重要依据。论文面向岩心含油荧光薄片图像,提出了一种基于图像分割和机器学习的微观剩余油分析方法。首先利用形态学开闭重建运算对图像的形态学梯度图像进行梯度重建,然后进行分水岭分割,应用区域合并算法对分割后的图像进行子区域合并,最后采用机器学习技术提取各区域的颜色特征并识别和分类。实验结果表明,本文提出的算法有效解决了分水岭算法的过分割问题,较好地实现了岩心中矿物颗粒分类识别和标注。

The remaining oil is mainly adsorbed in underground rocks.The key to solve the remaining oil problem is to study the carrier of the remaining oil,and the microscopic occurrence state of the remaining oil in the pores of the reservoir is an important basis to guide the potential exploitation of the remaining oil.In this paper,a micro-residual oil analysis method based on image segmentation and machine learning is proposed for the image of oil-bearing fluorescent flake in cores.Firstly,morphological open and close reconstruction is used to reconstruct the gradient of the morphological gradient image,then watershed segmentation is carried out,region merging algorithm is applied to merge the sub-regions of the segmented image,and finally machine learning technology is used to extract the color features of each region and identify and classify them.The experimental results show that the proposed algorithm can effectively solve the oversegmentation problem of the watershed algorithm,and well realize the classification and identification of mineral particles in cores.