A Median Regularized Level Set for Hierarchical Segmentation of SAR Images

摘要: An efficient strategy of image processing algorithms to deal with the speckle noise is incorporate data knowledge and models into them. In this letter, we introduce a hierarchical level set algorithm, which fast precise for multiregion segmentation synthetic aperture radar (SAR) images. Our algorithm performs curve regularization nonparametric median filter instead using curvature formulation, hence it reduces computation time. The proposed also replaces front propagation derivatives by morphological operations, finally, arithmetic-geometric distance measures contrast between regions controls segmentation. We conducted experiments on real SAR images modeled $\mathcal {G}_{I}^{0}$ distribution. performance evaluation two related methods comprises time based accuracy stochastic distance. Overall, our performed faster more both