Abstract:
Real-Time Whiteboard Capture System captures whiteboard lecture video in real time and extracts newly written pen strokes. Unlike other systems, it does not require special instruments such as electronic pen or whiteboard. The captured video images are classified into pen strokes on the whiteboard, foreground (person in front of the whiteboard) and Whiteboard Background Model (whiteboard without the text). The newly written pen strokes are extracted from current image frame and updated to the Whiteboard Background Model. Whiteboard Balancing is done where the colour of the whiteboard is corrected and made completely white. The captured video is processed at 8 frames per second (fps) instead of 25 fps that is output from the camera. Whiteboard balancing and reducing the processing frame rate provide a reduced bandwidth when transmitting whiteboard contents to remote meeting participants. This system can be used in conjunction with other teleconferencing tools such as voice conference and application sharing to improve online meeting experience. Keywords: Whiteboard, real-time, luminance, pixels, archiving, foreground, bandwidth.
