On the motion compensation within a down-conversion decoder

The most straightforward approach in obtaining a down-converted image seque nce is to decimate each frame after it has been fully decoded. To reduce me mory requirements and other costs incurred by this approach, a down-convers ion decoder would perform a decimation within the decoding loop. In this wa y, predictions are made from a low-resolution reference which has experienc ed considerable loss of information. Additionally, the predictions must be made from a set of motion vectors which correspond to the full-resolution i mage sequence. Given these conditions, it is desirable to optimize the perf ormance of the motion compensation process. In this paper we show that the optimal set of filters for performing the low-resolution motion compensatio n is dependent on the choice of down-conversion filter The motion compensat ion filters are determined as the optimal solution of a least squares probl em. This problem is formulated in the context of two general classes of dow n-conversion techniques: one which is dependent on a single block, and anot her which is dependent on multiple blocks. General solutions for each class of down-conversion are provided. To demonstrate the usefulness of these re sults, a sample set of motion compensation filters for each class of down-c onversion is calculated, and the results are incorporated into a low-resolu tion decoder. In comparison to a sub-optimal motion compensation scheme, th e optimal motion compensation filters realize a drastic reduction in the am ount of drift. Simulation results also reveal that the filters which were b ased on multiple block down-conversion can reduce the amount of prediction drift found in the single block down-conversion by as much as 35%. (C) 1998 SPIE and IS&T. [S1017-9909(98)00603-5].