摘要
提出了一种基于微血栓运动分析的微血管特征结构自动提取策略。提出用灰度梯度直方图统计来自动选阈的快速阈值值化算法，检测形态复杂的微血管图像边缘，抑制次要的微血管，采用低阈值双窗二次角点选择策略选取边缘曲线角点。通过微血管显微图像及其二值化图像分析，建立反映含微血栓的微血管特征结构模型，利用微血管的先验知识，给出提取微血管特征结构的算法，最后给出微血管显微图像结构的提取结果，实验证明该算法是十分有效的。含微血栓的微血管的特征结构建立，复杂的微血栓的匹配和识别问题将得到简化，微血管及微血栓的形态变化及运动估算任务得以减轻。该研究对于脑微循环障碍和老年病的基础医学研究和临床实践具有十分重要的意义。
A shrategy for automated extraction of the micro-vessel characterstic structure based on motion analysis of white micro-thrombus is presented in this paper.A fast automated choosing threshold algorithm using gray gradient-gray histogram statistics made into binary image is proved to be valid. After studying profoundly the forms of contour of non-rigid biomedital objects such as contour of micro-vessels, an algorithm of comer-point extraction of curves depicted by chain code which is so-called twice determining corner-point set with a low threshold and two window sizes is provided. Some achieved results show that the strategy is useful and helpful. To build the characteristic is useful and efficient for motion estimation of white micro-thrombus and calculation of the morphologic morphologic parameters of micro-vessel and white micro-thrombus. It is meaningful for studying cerebral micro-circulation obstacle and the veteran disease on clinic and basic medicine.
出处
《生物物理学报》
CAS
CSCD
北大核心
1996年第2期289-296,共8页
Acta Biophysica Sinica
基金
国家自然科学基金
关键词
血管
微血管
微血栓
特征结构
Edge detection
Chain-code curvature
Corner-point extraction White micro-thrombus
Micro-vessel
Characteristic structure
