APPARATUS AND METHOD FOR CONVERTING OBJECT-BASED VIDEO DATA INTO STEREOSCOPIC VIDEO

摘要

A device and a method for converting a 3D(Dimensional) image of object-based video data divided into a foreground/background object region are provided to generate the natural and stable 3D image efficiently by using the MPEG(Moving Picture Experts Group)-4 data, as an object motion type determiner determines/offers the type of diverse motions of the object-based video data to a 3D object generator. An image extractor(100) extracts image data from analog or digital data, or the MPEG-4 data. An object region segmenting part(110) segments the extracted image data into the foreground/background object region. An object motion vector generator(120) generates motion vectors respectively matched with blocks in a foreground/background object. The object motion type determiner(130) determines the motion type of the object matched with the current image by using the motion vectors. The 3D object generator(140) generates the different image matched with the current image by using the motion type and generates the 3D image from two images. A 3D object post-processor(150) fills an empty pixel of the 3D image with a proper color. The first and the second converter(160,170) convert the 3D image into an analog signal.