Automatic data capture for geographic information systems

This paper presents a map interpretation system for automatic extraction of high level information from the scanned images of Chinese land register ma ps. Our map interpretation system consists of three main components: text/g raphics separation, parcel extraction, and rotated character recognition. O ur approach to text/graphics separation is based on a simple yet effective rule: the feature points of characters are more compact than those of graph ics. In the parcel extraction process, the proposed algorithm traces the br anches between feature points to extract polygon structure from line drawin gs. Our character recognition method is based on the matching of extracted strokes using a neural network. The techniques of text/graphics separation and character recognition are robust to the rotation and writing style of c haracter. Another advantage of our separation algorithm is that it can succ essfully extract a character connected to graphical line. Experimental resu lts have shown that the proposed system is effective for the data capture o f geographic information systems.