A Novel Caption Insertion Method Based on DCT Domain
Jun Sun · Jisuanji gongcheng · 2004
This paper applies a novel compressed domain approach to insert caption in MPEG video stream. This approach reduces a lot of computing complexity and storage memory comparing to conventional method of decompression-addition-recompression. In order to avoid the presence of undesired semi-transparent region around the caption, a different approach from the usual pixel domain compositing operation is adopted.