METHOD FOR INTERPOLATING A PREVIOUS AND SUBSEQUENT IMAGE OF AN INPUT IMAGE SEQUENCE

摘要

Method for interpolating a previous and subsequent image of an input image sequence (SI), comprising the following steps: determining fidelity information (SF<SUB>I</SUB>) for at least one motion vector (SV<SUB>M</SUB>) which is descriptive for motion between said previous and subsequent image, wherein said fidelity information (SF<SUB>I</SUB>) is descriptive for the level and accuracy of said motion, determining classification information (SC<SUB>F</SUB>) for said at least one motion vector (SV<SUB>M</SUB>), wherein said classification information (SC<SUB>F</SUB>) depends on said fidelity information (SF<SUB>I</SUB>), said classification information (SC<SUB>F</SUB>) being descriptive for the motion type of said motion, and selecting an interpolation method in dependence of the determined classification information (SC<SUB>F</SUB>).