基于LBF改进模型的煤岩细观结构研究

Investigation on meso-structure of coal and rock based on the modified LBF model

为了准确全面地获取灰度不均匀的煤岩图像的细观结构信息,引进了基于LBF模型的图像处理方法,并针对其处理灰度不均一图像效率低和分割结果依赖于初始轮廓位置和大小的缺点,分别引入图像增强算子和规范初始水平集函数进行改进。通过对煤岩的3种不同图像进行灰度化、图像分割、形态学处理,获得了高质量的煤岩细观结构二值图像,并分析了各细观参数的统计规律。试验结果表明,基于LBF改进模型的图像处理方法能有效处理灰度不均匀的煤岩图像,获得的细观结构信息可为数值模拟煤岩试件提供真实的几何模型,为煤岩细观损伤力学分析提供可靠的技术支持。

For accurately and completely capturing the meso-structure information of coal and rock images which combine with the characteristics of inhomogeneity intensity extensively, the anthor introduces an image processing method based on Local Binary Fitting （LBF） energy model. Also, the modifications are carried out by introducing the image enhancement operators and regulating the initial set function for the defects of inefficiency in dealing with the low contrast images and the segmentation depending on the location and size of initial outline. The high-quality binary images of coal and rock meso-structure were obtained by graying, image segmentation and morphological processing for three kinds of images, and then analyzed the statistical regularity of the meso-parameters. The test results show that the im- age processing method based on modified LBF model can effectively deal with the uneven gray images of coal and rock, and the obtained meso-structure information can provide a true geometric model for coal and rock specimen nu- merical simulation and reliable technical support for meso-damage mechanics analysis.