Image processing apparatus and method of motion vector detection in a moving picture, and recording medium used therewith

An image processing method for detecting motion vectors in an image composed of moving picture data and constituted by at least one first pixel blocks that is constituted by more than one sub pixel blocks, comprising:

• a reduced-image generating step for generating a reduced image constituted by at least one second pixel block,
• wherein the second pixel block is generated by reducing the first pixel block with a predetermined factor;
• a first detecting step for detecting motion vectors of the second block in the reduced image as common vectors;
• a setting step for setting, by extending the common vectors by the predetermined factor, a plurality of possible motion vectors of said first pixel blocks;
• an extracting step for extracting, for a sub pixel block, second motion vectors already detected by a second detecting means for adjacent sub pixel blocks and possible motion vectors for adjacent sub pixel blocks in the same picture;
• a selecting step for selecting a base vector from the second motion vectors or possible motion vectors extracted in said extracting step;
• and
• a second detecting step for detecting, in a predetermined search range around the base vector selected in said selecting step, second motion vectors for the sub pixel block of interest in the image,
• wherein the size of the predetermined search range when the selected base vector is a second motion vector is different from the predetermined search range when the selected base vector is a possible motion vector.

A computer-readable medium having a computer-executable program recorded thereon, when executed, the program causes the computer to detect motion vectors in an image composed of moving picture data and constituted by at least one first pixel blocks that is constituted by more than one sub pixel blocks, comprising:

• a reduced-image generating step for generating a reduced image constituted by second pixel blocks smaller,
• wherein the second pixel block is generated by reducing the first pixel block with a predetermined factor;
• a first detecting step for detecting motion vectors of the second pixel blocks in the reduced image as common motion vectors;
• a setting step for setting, by extending the common motion vectors by the predetermined factor, a plurality of possible motion vectors of said first pixel;
• an extracting step for extracting, for a sub pixel block of interest in the image, second motion vectors already detected by a second detecting means for adjacent sub pixel blocks and possible motion vectors for adjacent sub pixel blocks in the same picture;
• a selecting step for selecting a base vector from the second motion vectors or possible motion vectors extracted in said extracting step;
• and
• a second detecting step for detecting, in a predetermined search range around the base vector selected in said selecting step, second motion vectors for the sub pixel block of interest in the image,
• wherein the size of the predetermined search range when the selected base vector is a second motion vector is different from the predetermined search range when the selected base vector is a possible motion vector.