Whiteboard scanning and image enhancement

A whiteboard can be an easy tool for collaboration such as brainstorming, and is widely used, but the content on a whiteboard is hard to archive and share. While digital cameras can be used to capture whiteboard content, the images are usually taken from an angle, resulting in undesired perspective distortion. They may contain other distracting regions such as walls and shadows. The visual quality of those images is usually poor. This paper describes a system that automatically locates the boundary of a whiteboard, crops out the whiteboard region, rectifies it into a rectangle, and corrects the color to make the whiteboard completely white. In case a single image is not enough (e.g., large whiteboard and low-resolution camera), we have developed a robust feature-based technique to automatically stitch multiple overlapping images. The system has been tested extensively, and very good results have been obtained.