基于超像素的图像智能算法在矿物颗粒分割中的应用

砂岩薄片岩石鉴定是油田勘探开发综合研究中认识储层特征、获取油气藏相关参数的重要基础工作;砂岩薄片鉴定的基础是矿物颗粒的分割,即把砂岩薄片图像中大量的矿物颗粒分割清楚.显微镜下砂岩薄片的图像场景复杂、矿物颗粒多,多前景目标是砂岩薄片图像分割的难点之一;不同矿物颗粒相似度高、边界模糊,传统的基于边缘特征的图像分割算法难以胜...

Application of intelligent image algorithm based on super-pixel in mineral particle segmentation

Sandstone thin section analysis is an important basis to understand the properties of reservoir and obtain the related parameters of oil and gas reservoir in the comprehensive study of oilfield exploration and development.Its basic work is to segment the mineral particles,that is,to segment the large number of mineral particles in the sandstone thin section image.Multi-foreground target is one of the difficulties in image segmentation since the image scene of sandstone thin section under microscope is complex with lots of mineral particles;due to the high similarity and fuzzy boundary of different mineral particles,the traditional image segmentation algorithm based on edge feature is not competent.To solve the above difficulties,this paper proposes a new intelligent segmentation method of mineral particles based on super-pixel features.The image is divided into a set of super pixels by comprehensive utilization of orthogonal polarized light and single polarized light image features.Then,the color,edge and texture features of the super-pixel are calculated,and the similarity of the super-pixel is further calculated to carry out regional fusion,so as to realize the mineral particle segmentation in the sandstone thin slice image,and reduce the effect of particle similarity and boundary blur on segmentation results.The experimental application shows that the accuracy of mineral particles and areas segmented by this method is more than 80%,which is obviously better than that of the traditional image segmentation algorithm.