POSITION INFORMATION ADDING APPARATUS, POSITION INFORMATION ADDING METHOD, AND COMPUTER PROGRAM FOR ADDING POSITION INFORMATION AND POSITION DETECTION APPARATUS

摘要

A position information adding apparatus includes: a reference position setting unit which sets a plurality of reference positions for each of a plurality of pictures contained in digitalized video data; a region setting unit which sets a first position information adding region at a first position defined based on at least any of the plurality of reference positions and a second position information adding region at a second position different from the first position, for each of the plurality of pictures; and a position information embedding unit which embeds a first moving pattern which moves on the pictures in a first temporal period into the first position information adding region and a second moving pattern which moves on the pictures in a second temporal period into the second position information adding region.

主权项

1. A position detection apparatus for detecting a reference position on a picture which is contained in digitized video data, and in which a first moving pattern which moves in a first temporal period is embedded on a first position information adding region having a first positional relationship with the reference position and a second moving pattern which moves in a second temporal period is embedded on a second position information adding region having a second positional relationship with the reference position, the position detection apparatus comprising:
a feature quantity extracting unit which obtains a feature quantity for each pixel of the picture which represents at least either a movement of a pattern captured on a region containing the pixel or a spatial feature of the pattern; an evaluation value calculating unit which obtains an evaluation value, for a predetermined region on the picture, representing a degree of likelihood that the predetermined region is at the reference position, based on the feature quantity of a pixel within a region having the first positional relationship with the predetermined region and the feature quantity of a pixel within a region having the second positional relationship with the predetermined region; and a reference position detection unit which detects the reference position based on the predetermined region with a maximum evaluation value.