基于微粒群算法的视网膜血管自动提取方法

摘要

In terms of the special gray distribution and region structure in retinal image, a novel blood vessel segmentation method based on Particle Swarm Optimization(PSO)and gray level-gradient co-occurrence matrix is proposed in this paper. The Gabor wavelet is taken to enhance vessels, then the maximum entropy of the gray level-gradient co-occurrence matrix optimizes and the best threshold value is received by PSO, after the retinal image’s binarization, the mathematical mor-phology is applied to extract the vessels region by analysis of region connectivity. The experiments implemented on the Hoover database indicate that the efficiency of the proposed approach, and this blood vessel segmentation method is bet-ter than the Hoover algorithm.%针对视网膜血管网络灰度分布特征与结构特征，提出了将灰度-梯度共生矩阵最大熵与微粒群算法相结合的视网膜血管提取方法。采用Gabor滤波以增强血管图像，获取增强后视网膜图像的灰度-梯度共生矩阵，利用微粒群算法并结合灰度-梯度共生矩阵的最大熵方法进行阈值化处理，对图像进行二值化处理后根据视网膜血管具有区域连通性的特征，采用形态学方法分割出最终的血管。实验结果表明，该方法能有效地提取视网膜血管网络。