Design new filter for linear image restoration

Abstract EN

A new restoration filter, which is based on combining Wiener filter and Constrained least squares filter were introduced.

The adopted filter was designed for restoring images that are degraded with space-invariant point spread function “PSP’ and corrupted with additive noise.

The results using an adaptive filter were compared, quantitatively, using root mean square error (RMSE), with those using conventional filters.

Different degradation parameters, i.e.

different blurring values and different signal to noise ratio were considered.

The result shows that better performances were obtained when using the adaptive filter compared with conventional filters.