煤岩显微组分中渗出沥青体角点特征检测方法

Corner Feature Detection of Exsudatinite in Macerals of Coal

渗出沥青体是煤岩显微结构中壳质组的重要组分之一。根据煤显微图像渗出沥青体具有的尖角特点,提出1种基于锐角特征的渗出沥青体检测方法。该方法在SUSAN算法的基础上改进而来,将SUSAN算法中模板圆的半径设置为r,动态改变r值,检测被模板圆覆盖的目标面积是否小于某一设定的阈值,如果是,则中心点判断为角点,同时估算出角点对应的角度数。检测结果表明,采用本文方法能够准确检测出渗出沥青体中所含的尖角,从而对相应组分作出判断。相比于其它方法,本文提出的方法算法简单、抗干扰性能较好。

Exsudatinite is an important component of exinite in coal microstructure. According to the sharp charac- teristics of exsudatinite, an exsudatinite detection method based on sharp angle feature is proposed. The method is modified from SUSAN algorithm, setting the SUSAN template circle radius to be r, increasing the r value gradual- ly, and detecting whether the target area covered by template circle is less than a threshold that is set in advance. If it is, then the center is determined to be the angular point, and the angle corresponding to the angular point is esti- mated. Experimental results indicate that the proposed method can accurately detect the angles contained in exsu- datinite, and the corresponding component is also judged. Compared with other methods, the proposed method is simpler and with good anti-noise performance.