Motion vector detection method and apparatus

An apparatus of forming an interpolation image at a time position of a (m+k)-th frame (k is an real number) between a m-th frame (m is an integer) and a (m+n)-th frame (n is an integer not less than i+l and i an integer not less than 1), the apparatus comprising:

• an extracting unit configured to extract a first block from a motion vector search area of a m-th frame;
• an extracting unit configured to extract, from a motion vector search area of a (m+n)-th frame, a plurality of second blocks selected in order of decreasing correlation with respect to the first block;
• a detecting unit configured to detect a plurality of first motion vectors between the first block and the plurality of second blocks;
• a calculating unit configured to calculate a plurality of second motion vectors which are i/n times the first motion vectors;
• an extracting unit configured to extract, from the (m+i)-th frame, a plurality of third blocks located at a destination of the first block according to the second motion vectors;
• a selecting unit configured to select, from the plurality of first motion vectors, as an optimum motion vector a motion vector corresponding to a maximum one of correlations between the first block and the third blocks.
• a scaling unit configured to scale the optimum motion vector according to a time position of the (m+k)-th frame;
• an extracting unit configured to extract, from the (m+i)-th frame, a fourth block located at a destination of an interpolation block located at same spatial position as that of the first block in the (m+k)-th frame, according to the scaled optimum motion vector;
• and
• an allocating unit configured to allocate the fourth block to the interpolation block located at the same spatial position as that of the first block to create the interpolation image.