定性与定量地描述冠状动脉血管，很大程度依赖于造影图像中的血管结构识别结果，对此，提出了一种多特征模糊识别算法判别血管结构．实现过程中，首先通过图像预处理获得血管初始特征，然后利用一圆周探测器沿血管路径扫描并获取多种局部测度；在定义各种局部测度的多特征模糊子集及其隶属度函数之后，通过构造模糊识别算子准确地判别血管的端、段、分支和交叉结构．该方法在仿真血管模型和多套实际冠状动脉造影图像上获得了较好的效果，对实际图像的结构识别平均正确率达到92 。-Qualitative and quantitative descr iption of coronary artery, a large extent depends on the contrast images of the vascular structure of the recognition result, this paper presents a fuzzy multi-feature recognition algorithms distinguish vascular structures. Implementation process, the first vessel through the initial characteristics of image preprocessing was, and then use a circular detector scanning along the vascular path and to obtain a variety of local measures　in defining the characteristics of the various local measure of the multi-fuzzy sets and membership function after by constructing a fuzzy recognition accurately determine vessel operator s side, section, branch and cross-structure. The method in simulation models and multi-vessel coronary angiography image sets were real good results, the structure of the actual image recognition accuracy of 92　on average.