Galaxy Image Restoration with Shape Constraint

Images acquired with a telescope are blurred and corrupted by noise. The blurring is usually modelled convolution the Point Spread Function noise Additive Gaussian Noise. Recovering observed image an ill-posed inverse problem. Sparse deconvolution well known to be efficient technique, leading optimized pixel Mean Square Errors, but without any guarantee that shapes of objects (e.g. galaxy images) contained in data will preserved. In this paper, we introduce new shape constraint exhibit its properties. By combining it standard sparse regularization wavelet domain, Shape COnstraint REstoration algorithm (SCORE), which performs deconvolution, while preserving shapes. We show through numerical experiments approach leads reduction ellipticitiy measurement errors at least 44%.