Motion Detection in Video Retrieval Using Content-Based Video Retrieval

The most important challenge in content-based video retrieval is detection of moving objects accurately. The videos are stored in the database and their features are extracted. The features of the input query clip were also extracted and compared with the database features. If the features are matched up to the threshold values then the similar videos are identified. This paper proposes a technique to detect the moving objects in the videos and in order to detect the objects an abstract difference technique is used. This technique is followed by noise filtering.