Contentbased Lecture Video Indexing Martin Halvorsen Contentbased Lecture

Content-based Lecture Video Indexing Martin Halvorsen Content-based Lecture Video Indexing

Content • • • Introduction Research Questions Background Proposed System Implementations Conclusions Content-based Lecture Video Indexing

Introduction Content-based Lecture Video Indexing

Introduction • Blackboard presentations considered as essential and indispensable • Hard to navigate through traditional videos • Lack of expertise and time consuming Content-based Lecture Video Indexing

Research questions Content-based Lecture Video Indexing

Research questions • Q 1: How can foreground/background segmentation in a lecture video work for different writing-boards? Content-based Lecture Video Indexing

Research questions • Q 2: How to automatically extract meta-data from lecture videos? Content-based Lecture Video Indexing

Research questions • Q 3: How to use such meta-data for indexing and searching of lecture videos? Content-based Lecture Video Indexing

Background Content-based Lecture Video Indexing

Background • Tracking, detection and separation algortihms • Motion estimation, background models, statistical estimated models Content-based Lecture Video Indexing

Background • Eirik Grythe – segmentation and text detection • Synne Repp – simulated indexing alg. Content-based Lecture Video Indexing

Proposed System Content-based Lecture Video Indexing

Proposed system • Superior blockdiagram of the proposed system F/B Segmentation Blackboard extraction Metadata Extraction Content-based Lecture Video Indexing Index Generation

Teacher segmentation Content-based Lecture Video Indexing

Teacher segmentation • Motion estimation - SAD Content-based Lecture Video Indexing

Teacher segmentation • Morphological operations to close holes in the detection Content-based Lecture Video Indexing

Teacher segmentation • Using motion history to improve the segmentation algorithm Content-based Lecture Video Indexing

Teacher segmentation • Replace blocks that has motion in it Content-based Lecture Video Indexing

Teacher segmentation • Demonstration Content-based Lecture Video Indexing

Meta-data extraction Content-based Lecture Video Indexing

Meta-data extraction • What is meta-data? • How to extract blackboard content? • Statistics Content-based Lecture Video Indexing

Meta-data extraction • Demonstration Content-based Lecture Video Indexing

Indexing Content-based Lecture Video Indexing

Indexing • When to extract an image? • Search a lecture video using extracted features Content-based Lecture Video Indexing

Conclusions Content-based Lecture Video Indexing

Conclusions • Adaptive segmentation possible by using motion estimation • Extraction of content is possible using image difference • Meta-data is possible to extract using a foreground model to compare against new content • Statistics of extracted meta-data can be used to automatically index a lecture video Content-based Lecture Video Indexing

The end Content-based Lecture Video Indexing