Intuitive interface for conceptual 3D modeling
MetadataShow full item record
The standard CAD tools allow great degree of control in 3D shape creation by requiring the user to specify exact shape parameters in a highly structured manner. However, that structure requires long training and familiarization and confines creative genesis, which flourishes more in open, interactive, and explorative environments. Currently, there is a lack of suitable methods to support ideation activities directly in 3D modeling software. In addition, the user interaction in current generation 3D modeling software is non-intuitive. Perhaps equally important, because of the dominant WIMP (windows, icons, menus, pointer) interaction mode, the world of highly structured 3D tools may largely be closed to the creative designer to begin with. The "steep learning curve" involved with WIMP mode of interaction to comprehend the complex task of using a dense set of toolbars and menus is not the best conduit to utilize creativity for a novice user. In this dissertation, I present an intuitive CAD framework for creating conceptual 3D models. I also discuss the construction of an ideation-modeling tool that enables novice users to rapidly construct conceptual 3D models from gestural interaction and suggested components from databases, with minimal training. Specifically, I developed (a) computational methods to extract high-level geometric, and functional aware abstractions from unorganized and heterogeneous 3D data and made it available for user interaction to support ideation 3D modeling, and (b) computational methods to enable intuitive gesture-based interactions in ideation 3D modeling. The developed computational methods were validated separately through the collection of testing data from a variety of users. The experimental user study results suggest that the developed framework is able to create variety of 3D conceptual models and overcome several limitations of the traditional CAD interfaces.