No abstract

No abstract

No abstract

No abstract

The benefits of the existing CAD systems are not available to a large number of engineering drawings which were created on paper using conventional drafting methods. There is a need for an automated system which will convert the scanned images of these drawings into a format recognizable by a CAD system. One of the main components of such a system is the vectorization (raster-to-vector) system which converts the binary image of the scanned drawing into a set of straight line segments, arcs, and circles. The accuracy of the vectorized output in terms of the location and size of the various entities is very important if it is to be used for further processing. This paper describes a system which rectifies errors in the vectorized output using a constraint-based approach based on the principles of variational geometry. The system considers drawings consisting of three orthographic views. The dimension sets present in the drawing and the logical relationship existing between the views are recognized and interpreted to obtain a set constraints governing the geometry of the object. The set of constraint equations are then numerically solved to obtain the exact coordinates of the points used to describe the geometry.