Maria Siebert

Lecture Video Indexing and Analysis Using Video OCR Technology

The text displayed in a lecture video is closely related to the lecture content. Therefore, it provides a valuable source for indexing and retrieving lecture video contents. Textual content can be detected, extracted and analyzed automatically by video OCR (Optical Character Recognition) techniques. In this paper, we present an approach for automated lecture video indexing based on video OCR technology: Firstly, we developed a novel video segmenter for an automated slide video structure analysis. Having adopted a localization and verification scheme, we perform text detection secondly. We employ SWT (stroke width transform) not only to remove false alarms from the text detection, but also to analyze the slide structure further. To recognize texts, a multi-hypotheses framework is adopted, that consists of multiple text segments, OCR, spell checking and result merging processes. Finally, we implemented a novel algorithm for slide structure analysis and extraction by using the geometrical information of detected text lines. The accuracy of the proposed approach is proven by evaluation.

Automatic Lecture Video Indexing Using Video OCR Technology

During the last years, digital lecture libraries and lecture video portals have become more and more popular. However, finding efficient methods for indexing multimedia still remains a challenging task. Since the text displayed in a lecture video is closely related to the lecture content, it provides a valuable source for indexing and retrieving lecture contents. In this paper, we present an approach for automatic lecture video indexing based on video OCR technology. We have developed a novel video segmenter for automated slide video structure analysis and a weighted DCT (discrete cosines transformation) based text detector. A dynamic image constrast/brightness adaption serves the purpose of enhancing the text image quality to make it processible by existing common OCR software. Time-based text occurence information as well as the analyzed text content are further used for indexing. We prove the accuracy of the proposed approach by evaluation.

Establishing an Expandable Architecture for a Tele-teaching Portal

The complexity of web applications becomes a big problem, when applications are growing. Especially if new algorithms should be tested and compared, it is important to be able to separate the newly implemented functions from the core system. It is not enough to use standard frameworks to handle this, but to plan a maintainable structure for the application. This paper describes how a maintainable structure can be planned and implemented by using the Django framework while trying to avoid a big programming overhead. The new architecture for the main features and its advantages and disadvantages is described together with an example for a rating plug-in. An outlook on the possibilities for future developments using this architecture is provided as well.

Realization of an Expandable Search Function for an E-Learning Web Portal

When providing a huge amount of content to users, the search engine is an important part of the platform. On most web sites it is the most important entry point to the data archive. That is why it is an interesting research topic. In this paper we present a solution how it is possible to make the search function expendable by using plug-ins. We show, why this is required in our web application, which is constantly growing through the creation of more searchable meta data. Therefore we show, how our requirements can be fulfilled and discuss the accompanying advantages and disadvantages.

Improving search in tele-lecturing: using folksonomies as trigger to query semantic datasets to extract additional metadata

Tele-teaching, where recorded lectures are streamed via the internet, was identified as the easiest adoptable method to produce large amounts of e-learning content. Due to the nature of the e-learning material produced, it is lacking searchable data and metadata. Social Web technologies have been identified as one way to overcome this problem, because metadata will be generated by users. But still, the amount of metadata generated that way is not sufficient and the contextual information is missing. This information can be extracted from the Semantic Web, especially from Linked Data initiatives. This paper describes a workflow that utilizes the metadata generated by users to trigger the query of semantic datasets. These datasets are providing additional metadata, that can be extracted via an interface, which is publicly available. Thereby it is possible to supply new search strategies and achieve an extension of the search space for multimedia e-learning data. An extension of an existing search functionality and the similarity detection is finally described.