改进的数学形态学开闭滤波算法在细胞图像预处理中的应用

Improved Mathematical Morphology Open-Closing Filter for Cell's Image Preprocessing

对数学形态学开闭滤波算法进行了改进,使其能针对癌细胞图像特征进行开闭滤波,在滤除背景噪声的同时,能滤除正常的离散细胞,保留疑似癌变细胞的全部原始图像特征.与其他几种滤波方法在癌细胞图像中应用的结果相比较,改进后的数学形态学开闭滤波算法能够很好地保留原始灰度信息,并且能最大限度地去除背景噪声及无用信息.

The mathematical morphology open-closing filter has been improved according to the characters of the tumor cell＇s image. The noise and the natural discrete cells in the image can be removed and the large cells and the cell groups can be reserved in the image. Comparing with some other filtering algorithms the improved mathematical morphology open-closing filter can reserve the most graylevel information of the original image,and remove the most of the useless information in image.