Content-Based Video Retrieval

Focusing on content-based video retrieval, this book explores innovative techniques to bridge the semantic gap between low-level video features and higher-level semantics. Dr. Milan Petković and Prof. Dr. Willem Jonker present three main techniques: a spatio-temporal extension of the Cobra framework, hidden Markov models, and Bayesian belief networks. They evaluate these methods through experiments with real video data and discuss the adaptability of their solutions across multiple domains. Additionally, a prototype video database management system is described, showcasing their proposed techniques.