VIDEO IMAGE PROCESSING

摘要

The appropriate one of a list of possible motion vectors is applied to each region of a video image. An image value for that region is determined for each of four successive fields, for each of the possible motion vectors. The values for the four fields are compared and when the values are substantially the same for a given vector it is assumed that that is the vector apppropriate to that region. The vectors appropriate to regions of obscured or revealed background can also be assigned by comparing inter-field differences and determining regions where the inter-field differences are high for the central pair and one outer pair and low for the other outer pair of fields. To generate regions corresponding to obscured background only information from the earlier fields is used and to generate regions corresponding to revealed background only information from the later fields is used.