PROJECT TITLE :
Structuring Lecture Videos by Automatic Projection Screen Localization and Analysis
We tend to present a absolutely automatic system for extracting the semantic structure of a typical educational presentation video, which captures the full presentation stage with abundant camera motions like panning, tilting, and zooming. Our system automatically detects and tracks each the projection screen and the presenter whenever they're visible within the video. By analyzing the image content of the tracked screen region, our system is ready to detect slide progressions and extract a high-quality, non-occluded, geometrically-compensated image for each slide, resulting in a list of representative pictures that reconstruct the most presentation structure. Afterwards, our system acknowledges text content and extracts keywords from the slides, which will be used for keyword-primarily based video retrieval and browsing. Experimental results show that our system is able to generate a lot of stable and accurate screen localization results than commonly-used object tracking ways. Our system also extracts a lot of accurate presentation structures than general video summarization ways, for this specific type of video.
Did you like this research project?
To get this research project Guidelines, Training and Code... Click Here