关键词: 各向异性扩散, 脉冲发放皮层模型, 合成孔径雷达图像分割, 边缘检测, 相干斑

Abstract:

An improved speckle reducing anisotropic diffusion (SRADPRO) algorithm was adopted to reduce the speckle in synthetic aperture radar (SAR) image. An adaptive SAR image segmentation algorithm speckle reducing spiking cortical model (SRSCM) was proposed when SRADPRO and spiking cortical model (SCM) was combined. In SRSCM, the standard deviation was calculated through a uniform sample region in the SAR image, and then the effect degree of speckle can be estimated as a result whether to employ the operator of SRADPRO was determined. At the second stage, the SCM operator was executed. Because of the auto wave characteristic of SCM, the segmented image and edge detection result can be obtained together. The effectiveness of SRSCM was proved by the comparisons with several traditional algorithms.

Key words: anisotropic diffusion, spiking cortical model, synthetic aperture radar image segmentation, edge detection, speckle