An Effective CBVR System Based on Motion, Quantized Color and Edge Density Features

Rapid development of the multimedia and the associated technologies urge the processing of a huge database of video clips. The processing efficiency lies on the search methodologies utilized in the video processing system. Usage of inappropriate search methodologies may make the processing system ineffective. Hence, an effective video retrieval system is an essential pre-requisite for searching a relevant video from a huge collection of videos. In this paper, an effective content based video retrieval system based on some dominant features such as motion, color and edge is proposed. The system is comprised of two stages, namely, feature extraction and retrieval of similar video clips for the given query clip. Prior to perform the feature extraction, the database video clips are segmented into different shots. In the feature extraction, firstly, the motion feature is extracted using Squared Euclidean distance. Secondly, color feature is extracted based on color quantization. Thirdly, edge density feature is extracted for the objects present in the database video clips. When a video clip is queried in the system, the second stage of the system retrieves a given number of video clips from the database that are similar to the query clip. The retrieval is performed based on the Latent Semantic Indexing, which measures the similarity between the database video clips and the query clip. The system is evaluated using the video clips of format MPEG-2 and then precision-recall is determined for the test clip.