Knowledge-Based Contour Map Recognition

Period of Performance: 01/01/1994 - 12/31/1994


Phase 1 SBIR

Recipient Firm

Elcee Computek, Inc.
500 NE Spanish River Blvd,, Suite 105D
Boca Raton, FL 33431
Principal Investigator


Elcee Computek, Inc. proposes a novel approach to develop a contour map recognition software application program. This application entails the design and implementation of a automated contour map conversion software program called Knowledge-based Contour Map Recognition (KCMR). The proposed system will take advantage of state-of-the-art image processing methods to (1) extract texts, symbols and then curves from images of contour line drawings, and to (2) identify and validate objects suitable for contour maps, based on prevailing knowledge. Phase I will identify the requirements to achieve the automated contour map object recognition algorithm and will demonstrate the procedure with a sample model/drawing. It will include the detailed study of candidate techniques and the specification of a proposed system architecture, adaptation and refinement of known methods, and the delivery of, a prototype software system. The KCMR Phase I prototype will utilize modem Graphical User Interface methods to provide a coherent and easy-to-use environment, complete with pull-down menus. Phase II refine the prototype's capabilities. This easy to use tool will provide for the transparent conversion of bitmapped data files into their equivalent GIS/CAD vector object database consistent with the Spatial Data Transfer Standard's (A K.A. FIPS 173) Conceptual Model of Spatial Data.