A method for scanning a document includes: acquiring scanned image data from a first region of the document; determining an initial estimate of a document attribute using pixels in the first region; acquiring scanned image data from a second region of the document; processing pixels in the second region in accordance with the initial estimate; determining a second estimate of the document attribute using selected pixels in the second region; determining if the initial estimate is valid and if not, processing pixels within the image in accordance with the second estimate.