Ee368 Project Title: Whiteboard Detection and Tracking of Presenter in a Video Lecture

Motivation Online video lectures are a very effective way to revisit the material after having watching live, or to have access to the lecture when impossible to attend it. A very common occurrence, both live and in video, is that the lecturer will stand in front of the whiteboard when trying to explain a concept (Figure 1). There is little that one can do being live in the audience, except maybe to change seat. However, having access to the video frames provides the possibility of removing the lecturer in order to see the material being presented. This problem has been addressed in [1] and [2]. We propose a simpler solution where we can track the presenter and its translucency can be adjusted, even in low contrast situations [3].