Blind restoration algorithm of total variation based on anisotropic regularizations
Tianxu Zhang · Infrared and Laser Engineering · 2007
A blind restoration algorithm of total variation based on anisotropic and nonlinear regularizations is proposed for restoring turbulence-degraded images,in which the anisotropic and adaptive regularizations are adopted according to the properties of turbulence Point Spread Functions(PSFs) and image.The nonlinear and spatial anisotropic regularization functions are suggested to smooth adaptively the gradients in the process of estimating the PSF and recovering the object image.Finally,the cost functions are minimized by alternate minimization scheme,and the nonlinear equations are lineared by fixed-point iteration scheme,then the PSFs can be estimated and the images can be recovered quickly.The robustness and effectiveness of the proposed algorithm are tested by a series of restoration experiments to recover the digital simulation and real turbulence-degraded images.