Creating Slides from Video Lecture

Abstract

Introduction. Video recordings of lectures are no longer a rarity in the conditions of distance learning. Videos may be in an inconvenient format for students or contain different artifacts due to compression, camera quality, and other factors. It is useful to have a presentation of the study material, which contains only the text from the board because such a view of the material is most like the compendium. Purpose. The item of our work is creating an algorithm for obtaining panorama slides without a teacher from video lecture. To develop this algorithm, we use the Boykov-Kolmogorov Max-flow algorithm for obtaining a mask of moving objects. In some video lectures, the camera is moved so that the teacher is in sight, so some of the recordings are not visible. To do this, we make frames stitching to get a panorama. To reduce duplication of slides due to camera shaking, we perform video-stabilization as a preprocessing step, then replace the pixels using the mask. Finally, we create slides by comparing changes in the frames, binarizingand denoising. Methods. We used the Boykov-Kolmogorov Max-flow algorithm, SIFT, Laplace operator and Otsu binarization for developing information technology. Results.As a result, we get panorama slides with the extracted text or drawings from the board, which will help to simplify the creation of e-learning materials for both new and existing lecture recordings. Teachers will also be able to quickly provide lecture material to students even if they teach several complex subjects. Conclusion. We have developed an algorithm and implemented information technology for obtaining slides from video-lecture. Now, a teacher can give students short content of video material. The next steps are automatic detection of the board, division of the board into sections, and recognition of the text and formulas written on the board.