煤与矸石分形维数的差异研究

发布时间：2018-03-11 15:08

本文选题：煤　切入点：矸石　出处：《煤炭科学技术》2017年07期 　论文类型：期刊论文

【摘要】：为实现煤矸自动识别及快速分选,利用煤与矸石的纹理特征与轮廓表面灰度级别等差异较大的特点,对于二者的单一颜色灰度图像,采取平滑滤波、降噪等方法处理,应用差分盒维法求得二者灰度图像分形维数,以区分煤与矸石图像表面纹理粗糙度。试验结果表明,煤和矸石的分形维数有明显的不同,煤的分形维数要大于矸石,以数值2.450 0为分界点,煤的分形维数在分界点之上,矸石的分形维数在分界点之下。根据这一差异可有望实现煤矸的自动识别,为实现一种全新的煤矸分选系统奠定理论基础。
[Abstract]:In order to realize the automatic recognition and fast sorting of coal and gangue, the texture features of coal and gangue and the grayscale level of contour surface are quite different. For the single color grayscale image of coal and gangue, the methods of smoothing filtering and noise reduction are adopted. In order to distinguish the surface roughness of coal and gangue images, the fractal dimension of grayscale image is obtained by differential box dimension method. The experimental results show that the fractal dimension of coal and gangue is obviously different, and the fractal dimension of coal is larger than that of gangue. The fractal dimension of coal is above the boundary point, and the fractal dimension of gangue is below the boundary point. According to this difference, it is possible to realize the automatic recognition of coal and gangue, which will lay a theoretical foundation for the realization of a new separation system of coal and gangue.
【作者单位】：中国矿业大学(北京)机电与信息工程学院;
【分类号】：TD94

【参考文献】