From Computer Vision to Document Recognition or Using Labeling Technique for Map Interpretation.
Sergey Ablameyko, O. Frantskevich · Digital Library of the Belarusian State University (Belarusian State University) · 1994
This paper considers an application of computer vision notions for document recognition problem. The well developed in computer vision labeling technique is applied for interpretation of black-and-white layers of geographical maps. The knowledge base and main relations between cartographical objects and segments are extracted and described. Map-drawing interpretation process by using labeling technique is proposed. The suggested approach allows to increase a level of automatic object recognition