多形核白细胞显微图像分析

为改变目前淋病病原体诊断存在的检验周期长，误诊率高的落后状况，将图像处理技术用于淋病病原体中多形核白细胞显微图像的分析．在对其进行了二维OTSU分割的基础上，通过4-邻域标识方法标记多形核白细胞区域，给出了自动测量细胞的面积、周长及形状因子等几何参数的有效方法．实验结果表明，得到的几何参数反映了多形核白细胞的形状特征，这为深入研究淋病病原体显微图像中淋球菌的分布创造了条件．该方法同样适用于其他类似传染病的显微镜分析处理．

Analysis of Microscopic Images of Multi-shape-nucleus Leukocyte

At present, the diagnosis of gonorrhea pathogen is time-consuming with a high error ratio. In order to overcome these shortcomings, this paper applies the technique of image processing to the measurement and analysis of the microscopic image of multi-shape-nucleus leukocyte. We propose a method, which first segments the image using the Two-Dimensional OTSU thresholding method, then identifies the multi-shape-nucleus region by the 4-neighbors mark method, finally actualizes the auto-measurement of multi-shape-nucleus leukocyte' s parameters such as dimension, perimeter, shape factor and so on. Experiments show that the determined parameters can well present the shape feature of the multi-shape-nucleus, which provides the prerequisite for further researching the distributing of the pathogen of gonorrhea; it also turns out to be suitable for microscopic image analysis of other similar epidemic.