Automatic Equalization Method for Synthetic Aperture Sonar Image Based on Background Estimation
To solve the problems of non-uniform grayscale,low contrast,and difficult discrimination of synthetic aperture so-nar(SAS)images,a method for equalization based on combined with local neighborhood and distance direction of image to adaptive background estimation is proposed.First,an equalization model based on the illumination model is proposed.Then,the non-uni-form time evolution model and equalization function is used to estimate the background component of the image through the differen-tial calculation.In the iteration process,the parameters are automatically optimized and judged to terminate according to the residu-al energy size.Finally,the equalization result is obtained through the illumination model and brightness parameter.Through actual data verification and analysis,it is found that the background in the SAS image becomes more even,the contrast is improved,and the targets and textures are enhanced,while noise and other interference signals are effectively suppressed after equalization.The al-gorithm performs excellently in terms of local equalization,information capacity,detail richness,peak signal-to-noise ratio,and structural similarity,which further verifies the practicality and effectiveness of the algorithm.
non-uniform grayscaleillumination modeltime evolution modelresidual energy
国家科技期刊平台
NSTL
万方数据
