A conceptual design tool: A sketch and fuzzy logic based system

Abstract

A real-time sketch and fuzzy logic based prototype system for conceptual design has been developed. This system comprises four phases. In the first one, the system accepts the input of online free-hand sketches, and segments them into meaningful parts by using fuzzy knowledge to detect corners and inflection points on the sketched curves. The fuzzy knowledge is applied to capture user's drawing intention in terms of sketching position, direction, speed and acceleration. During the second phase, each segmented subpart (curve) can be classified and identified as one of the following two-dimensional primitives: straight lines, circles, circular arcs, ellipses, elliptical arcs or B-spline curves. Then, two-dimensional topology information (connectivity, unitary constraints and pairwise constraints) is extracted dynamically from the identified two-dimensional primitives. From the extracted information, more accurate two-dimensional geometry can be built up by a two-dimensional geometric constraint solver. The two-dimensional topology and geometry information is then employed to further interpretation of a three-dimensional geometry. The system can not only accept sketched input but also users’ interactive input of two- and three-dimensional primitives. This makes it friendly and easier to use, in comparison with ‘sketched input only’ or ‘interactive input only’ systems. Finally, examples are given to illustrate the system.