提出了一种基于Hessian特征的视网膜血管图像的增强滤波算法,分析了Hessian特征值对各类形状的抑制和加强作用,将Hessian矩阵的特征值与特征向量应用于视网膜血管特征的响应函数及形态学和非线性扩散处理的各个环节,并选择合理的尺度空间范围和尺度空间增量,调节因子而平滑非线状区域和锐化增强血管区域,本文算法在同等准确率下具有较高的稳定鲁棒性。
This paper presents an enhancement filter algorithm of retina blood vessels based on Hessian matrix features.It firstly analyzes the effects of suppressing and enhancing of Hessian matrix eigenvalue on all kinds of shapes.Then it exploits these eigenvalues and vectors in the subsequent processing steps such as response function,mathematical morphology and nonlinear anisotropic diffusion to enhance whole blood vessel domain and smooth the nonlinear region.The proposed algorithm has higher stability robustness in the same accuracy.