Conversion from interlaced to progressive formats by means of motion compensation based techniques

The paper first shows how to improve the classical gradient method for HDTV in order to estimate the motion between two fields of the same parity. The improvement is obtained by means of a symmetric scanning of the picture and the estimation of the motion for small blocks rather than for isolated pels. By means of the estimated motion information, it then develops a prediction method for the pels which are missing in the field to be deinterlaced.<
>