RESEARCH ON AUTOMATIC THRESHOLD GENERATION METHOD FOR SHALE SLICE HOLE SEGMENTATION

With the rapid progress of shale gas exploration and exploitation, microstructural characteristics and their analysis techniques have become more and more important. Among them, to acquire the shale pore structure parameters based on the threshold segmentation method is an key mean for shale microstructural characterization. The existing methods mainly include the maximum between-class variance method and the maximum entropy threshold segmentation method. They have gained good results in various image segmentation. However, they are all quite time-consuming and cannot effectively separate the pores from matrix in the SEM images of thin slices sometimes. In this paper, an automatic threshold generation method for shale slices and fissures segmentation is proposed, which can adaptively and quickly generate the optimal grayscale threshold for the related image according to different characteristics of shale slices, and automatically recognize shale pores and matrix as well as other geological elements. In this paper, experiments were conducted for the shale scanning electron microscopy images from the Well 206, and comparison is made with the results from traditional methods. The experimental results show that the algorithm of this paper can automatically generate the optimal grayscale threshold on various images, and accurately identify such elements as pores and matrix. Therefore, the automatic threshold generation method for shale slice hole segmentation introduced in this paper can efficiently generate the optimal gray threshold of shale slice scanning electron microscope image and provide a reliable basis for quantitative analysis of pore structure on shale microscopic images.