论文部分内容阅读
为实现核孔膜防伪的自动检测,提出了一种基于数学形态学的核孔膜显微图像检测方法。构造了基于灰度数学形态学的二值化方法,提出了基于二值数学形态学的分层提取方法,统计总孔数和孔半径。实验结果表明,二值化方法不但能够有效消除灰度图像中的噪声和渐变背景,而且所得二值图像保留了原始灰度图像中核孔膜的全局信息;分层提取方法所得总孔数的相对误差在10%以内,同时实现了孔半径的统计。该方法可用于仪器化检测中。
In order to realize the automatic detection of nuclear membrane anti-counterfeiting, a detection method of nuclear membrane microscopic images based on mathematical morphology was proposed. A binarization method based on gray mathematical morphology was constructed and a hierarchical extraction method based on binary mathematical morphology was proposed. The total number of holes and the hole radius were calculated. Experimental results show that the binarization method not only can effectively eliminate the noise and the gradient background in the grayscale image, but also retains the global information of the nuclear pore film in the original grayscale image. The total number of holes obtained by the hierarchical extraction method is relatively The error is within 10%, at the same time realize the statistics of the hole radius. This method can be used in instrumental testing.