Content based video retrieval using support vector machine classification

Softwares are used to model, replicate and more importantly exceed human vision using different software technology. There is an exponential increase in the digital content which is being generated rapidly, particularly videos. Content-Based Video Retrieval (CBVR) is necessary for efficient search of these media. Still image retrieval were used initially as the content based visual queries. The capturing and encoding of digital media has resulted in high volume of visual information to be produced and spread rapidly is a challenging issue for recovering the videos. The proposed work has two parts training and classification to retrieve video from database based on query input from the user. To address this issue a CBVR framework is developed using MATLAB R2010a.