A method and device for capturing a positionally corrected image of a document is disclosed. The method comprises the steps of: obtaining a two-dimensional image of the document with a mobile terminal apparatus; performing edge detection within the two-dimensional image to identify edges of the document; determining angles between detected edges; calculating, based on the detected edges and the angles determined, a three-dimensional position of the document relative to a position of the mobile terminal apparatus; calculating correction information to correct, by relative movement, the position of the document relative to the position of the mobile terminal apparatus; providing first guidance information derived from the correction information to a user of the mobile terminal apparatus, guiding the user to perform the relative movement; and capturing a positionally corrected image of the document. Thereby, the document can be captured with a quality sufficient to permit Optical Character Recognition (OCR).