目的：研制快速、精确的视网膜血管中轴线（ｖｅｓｓｅｌａｘｉａｌｓｋｅｌｅｔｏｎＶＡＳ）提取算法，以定量分析眼底血管形态改变的细微变化。方法：１．图像预处理（去噪、增强）。２．以经二值化及膨胀后的梯度图像为模板，进行基于高斯匹配滤波法的快速血管提取。３．后处理（细化等）。结果：分别采用本文方法及传统方法对视网膜ＶＡＳ进行提取与结果比较。结论：本文方法具有快速、精确、抗噪声能力强及提取出的ＶＡＳ连通性好的优点 OBJECTIVE: To develop a fast and accurate extraction algorithm for vesselaxial keratometry VAS to quantitatively analyze the subtle changes in the vascular morphology of the fundus. Methods: 1. Image preprocessing (denoising, enhancement). 2. The rapid blood vessel extraction based on Gaussian matched filtering method was performed with the binarized and expanded gradient images as a template. 3. Post-processing (refinement, etc.). Results: The retina VAS was extracted and compared with the results of the method and the traditional method respectively. Conclusion: This method has the advantages of fast, accurate, strong anti-noise and good VAS connectivity