Edge-sensitive image restoration using order-constrained least squares methods

摘要: In this paper we consider a novel technique for the restoration of noise corrupted images, using order-constrained least squares methods. The process uses moving cross-shaped filter window, within which two operations are combined. first operation consists simple hypothesis tests presence an edge some minimal height δ, crossing center window. second computes window output as fit windowed values (if is deemed to be present), or simply average no present). new applied actual and results compared applying similarly configured median averaging filters. Some computational considerations comparisons discussed at end paper.