| 摘要 |
A digital image signal scaling method is provided to prevent the generation of a motion estimation error when enlarging a super high resolution image. When an interlaced scanning SD(Standard Definition) image is converted into a progressive scanning SD image through the de-interlacing, the size of converted image is the same as the HD(High Definition) image through a wavelet/bicubic image interpolating member(10). The converted image is used as an initial value of a parallel wavelet-POCS(Projection Onto Convex Sets) image enlargement member(20). An HD-SD motion estimation member(50) estimates motion by means of the present SD image and the HD image previously stored in an HD frame memory member(40).
|
