DUPLICATE VIDEO DETECTION APPARATUS AND ITS METHOD, RECORDING MEDIUM HAVING COMPUTER PROGRAM RECORDED

摘要

The present invention relates to a technique for determining whether a duplicate video exists, and the duplicate video detection apparatus divides the input video into shot units, extracts key frames from each of the divided shots, and then visualizes each extracted key frame. After extracting the information, the semantic information is detected using any one of the extracted visual information, a predefined semantic model, an image database and a Foxsonomy image, and then using the detected semantic information to generate a semantic based video index. By comparing the generated semantic-based video index with the reference video index, it is possible to easily determine whether the input video is duplicated.