| Feature identification from CAD model is an essential requirement for integrated design and manufacturing systems. Mainly there are two approaches to identify features. One is recognizing features directly from three-dimensional geometric description of a design model. The other is done using two-dimensional projections of the 3D model. In many cases, feature identification using two-dimensional projections can be more efficient than the three-dimensional approach. Some level of checking and correcting the input 2D CAD drawing is necessary to perform feature extractions. This research investigates the automated checking of 2D CAD drawing also called 2-D healing, and presents a methodology for automatically detecting various errors on a given drawing. For the downstream applications, it is essential to extract the feature semantics. The resultant operational procedure can ensure minimum number of setup changes and tool changes during the manufacturing process. Hence, the needed method is expected to extract semantic information on features. In particular, this research determines the semantics of a hole feature using spatial geometric reasoning applied to the 2-D data. New and robust algorithms are developed for determining the semantics on a hole feature. Results based on the algorithms are presented. |
