Nonstandard Diffusion in Image Restoration

We present new models for image restoration which arise from minimizing 
functionals with nonstandard growth conditions.   These models effectively 
remove noise while preserving and enhancing edges in degraded images.  In 
particular, they drastically reduce the 'staircasing effect' which can arise 
when reconstructing piecewise smooth images.  We present experimental results 
which demonstrate the effectiveness of the models in image restoration.  The 
existence, uniqueness and asymptotic behavior of the models are also 
discussed.