作　　者： ; ; ;

机构地区： 福州大学数学与计算机科学学院

出　　处： 《中国体视学与图像分析》 2004年第4期193-198,共6页

摘　　要： 目的:介绍了一种自动提取淋巴瘤病理图像中空泡状细胞核方法。方法:以空泡状细胞核为目标,采用在CMYK模式下的多层分割法分割病理图像;提出变尺寸特征窗口模糊分类法,对空泡状细胞核进行识别提取;并用近似最短连接法对破碎的空泡核进行修补连接。结论:实验结果表明,该方法可以有效对淋巴瘤病理图像进行分割,较准确地提取空泡状细胞核,并具有一定程度的自动修补连接破碎细胞核能力。 Objective: A new method is introduced to extract vacuole nucleus in the lymphoma pathologic image automatically. Method: With vacuole nucleus as the subject, use multilayer segmentation algorithm to segment pathologic images in the CMYK mode. Use a fuzzy classification with dynamic size feature windows to recognize and extract vacuole nucleus. In the meantime, restore and connect broken vacuole nucleuses with approximately nearest connection algorithm. Conclusion: The results of these experiments show that this method can effectively segment lymphoma pathologic images, comparatively accurately extract vacuole nucleuses, and, to some degree, be able to restore and connect broken nucleuses automatically.

关 键 词： 空泡状细胞核 模式 分割 模糊分类 复原

领　　域： [自动化与计算机技术] [自动化与计算机技术]