Weighted fuzzy filter on interlaced-to-progressive conversion

Abstract

In this paper, we describe a new spatial domain deinterlacing algorithm which based on the fuzzy weight function. This technique is applied to the issue of missing pixel calculation in video deinterlacing. Traditional edge-based deinterlacing methods became poor at detecting precise edge direction which influences deinterlacing performance. Also, they do not use all edge direction information together but just use only one edge direction. In the proposed algorithm, weights are extracted from luminance difference values and multiplied by the candidate deinterlaced pixels. The extracted weights are worth to be used for proposed method to calculate the edge direction precisely. Compared to the conventional edge-based deinterlacing method, the proposed algorithm provides good visual results in various kinds of edges.