A system enables users at different locations (e.g. different geographic locations) to work collaboratively on a document, and to view each other's annotations to the document. Each user can work with a paper document, and make handwritten annotations to the paper document. The system includes a plurality of workstations communicating with a server. Each workstation comprises a document-imaging device, such as a camera, for generating a digital image of the document. Handwritten annotations are identified by image processing, and new annotations made by each user can be tracked in real time. Image bitmaps of the annotations are distributed by the server to each workstation so that each user can view other user's annotations. In one view, the annotations from different users are superimposed in a combined image.