FIELD: physics.
SUBSTANCE: contours in the digital image are determined, the detected contours are divided into four sets of contours corresponding to four different sides in the original digital image, hypotheses about the edges or boundaries of the fragment of the digital image containing the document are constructed based on these contours and hypotheses are evaluated for choosing hypotheses with the highest estimates for representing the boundaries of the fragment of the digital image containing a document in the original digital image obtained.
EFFECT: enabling the detection of document fragments of a digital image.
22 cl, 48 dwg
