中文摘要：

该文提出了一种基于数学形态学理论的方法,用于分割细胞图像。首先利用灰度形态学中的重构运算对输入图像进行滤波来削弱或去除噪声的影响,然后取其形态学梯度并结合Top-Hat和Bottom-Hat两个变换来增强梯度图像的对比度,最终由分水岭算法完成分割。关于过分割的改进,该文引出一个特殊的Otsu阈值来标记图像,得到的结果无需进一步的后处理。

英文摘要：

A method is proposed in this paper which is based on the theory of mathematical morphology and used to segment cell images. The input image is first filtered using reconstruction of gray - scale morphology for the suppression or removal of noise effect, then turned into the image of morphological gradient, whose contrast is enhanced by combining the Top - Hat and the Bottom- Hat transforms, and finally segmented through the watershed algorithm. As for the improvement of oversegmentation, a specific Otsu threshold is introduced to mark the images and thus the obtained results need no further postprocessing.