Searching for Classes of Visual Content in Electronic Lectures

Lecture recording allows the creation of electronic learning material for local and distance education. Flexible screen recording techniques can capture virtually any material displayed during a presentation. Unlike other screen recorders, our TeleTeachingTool already offers slide-based navigation and full text search as meaningful retrieval features. Our paper describes how to locate remembered content, such as images, diagrams, or graphs, and furthermore how to find similar passages and related content in large databases of recorded lectures by classifying navigational indices. We introduce a classification scheme using characteristic parameters based on the analysis of color histograms of the recorded screens.