This paper explores the use of camera phones with pen input as a platform for generating digital annotations to realworld objects. We analyze the client-side requirements for a general annotation system that is applicable in mobile as well as stationary settings. We outline ways to create and interact with digital annotations using the camera and pen-based input. Two prototypically implemented annotation techniques are presented. The first technique uses visual markers for digital annotations of individual items in printed photos. The second technique addresses the annotation of street signs and indication panels. It is based on image matching supported by interactively established 4-point correspondences.