Assigning attributes to an image includes determining a particular capturing device used to obtain the image and detecting attributes of the image based on image data, where an expected location and content of the image data varies according to a particular set of rules that is chosen based on the particular capturing device used and/or a type of document represented by the image. The capturing device may be a scanner, a facsimile device, or a camera. The image may have an expected form factor. Some of the attributes of the image may be fully recognized and some of the attributes may be partially recognized and, for at least some of the partially recognized attributes, a lookup table may be used to clarify unrecognized portions thereof. A user may be prompted to facilitate identifying partially recognized attributes. Partially recognized attributes may be indicated to a user with question marks.