A Color-Content based Video Retrieval System with a Time-Sequence Analyzer

Ha Bui Hong and Yasushi Kiyoki

Keywords

Multimedia database, Video Retrieval, Time-Analyzer

Abstract

“Color Content Based Video Retrieval System with a Time Sequence Analyzer” (CCBIR-TSA) is a video retrieval system based on color features and supported by a time-sequence analyzer. In our system, the query is a sample video or a sequence of sample images, and the result is the video that has scenes with the similar color features. Our time-sequence analyzer guarantees that the resulting video has similar scenes, which appear in the same order as the sample scenes in the query. In order to implement a time-sequence analyzer, our system extracts color feature vectors of each frame in the videos and transforms video frames into comparable sequences of color feature vectors. Traditional “Longest Common Subsequence” algorithm (LCS) is applied to calculate the longest match between extracted frame sequences and video data content.

Important Links:



Go Back