Spatial de-interlacing using dynamic time warping

Spatial de-interlacing is an essential part of motion adaptive de-interlacing used for reconstructing missing lines in cases of fast motion detection. Common spatial de-interlacing algorithms often produce artifacts in the output image, especially along edges with flat horizontal angles. In this paper we introduce a new method for spatial de-interlacing based on the dynamic time warping (DTW) procedure. The DTW algorithm finds an alignment between two original consecutive lines, and then, the missing line between them is reconstructed based on this alignment. This method preserves the smoothness of the original image edges and produces a high quality progressive image.