A Novel Algorithm for Video Retrieval using Video Metadata Information

This paper proposed a video metadata information based method for video retrieval. Extracting key frames and forming video abstract first, and then segmenting the key frames into two parts—background and theme, calculating their gray-scale distribution and probability density, and then making up video metadata, and finally retrieving information through structuring video metadata. Experiment shows that this retrieval algorithm is efficient that people can search the information they want exactly and speedily through this method.