A method for area identification includes identifying a plurality of candidate predefined edges in a document image of a document. The candidate predefined edges are edges in a predefined direction of the document. The method further includes determining one of the candidate predefined edges to be a target predefined edge and identifying at least one information area in the document image based on the target predefined edge.