0
Research Papers

Shape Design From Exemplar Sketches Using Graph-Based Sketch Analysis

[+] Author and Article Information
Günay Orbay

Department of Mechanical Engineering,  Carnegie Mellon University, Pittsburgh, PA 15213

Levent Burak Kara1

Department of Mechanical Engineering,  Carnegie Mellon University, Pittsburgh, PA 15213lkara@cmu.edu

The size of the canonical form can be at most the size of the largest common subgraph.

Note that one of the feature vectors will be formed by identity transformation matrices as the reference shape and one of the deformed shapes are identical.

1

Corresponding author.

J. Mech. Des 134(11), 111002 (Oct 02, 2012) (14 pages) doi:10.1115/1.4007147 History: Received May 27, 2011; Revised July 01, 2012; Published October 02, 2012; Online October 02, 2012

We describe a new technique that works from a set of concept sketches to support the exploration and engineering of products. Our approach allows the capture and reuse of geometric shape information contained in concept sketches, as a means to generate solutions that can concurrently satisfy aesthetic and functional requirements. At the heart of our approach is a graph-based representation of sketches that allows the determination of topological and geometric similarities in the input sketches. This analysis, when combined with a geometric deformation analysis, results in a design space from which new shapes can be synthesized, or a developing design can be optimized to satisfy prescribed objectives. Moreover, it facilitates a sketch-based, interactive editing of existing designs that preserves the shape characteristics captured in the design space. A key advantage of the proposed method is that shape features common to all sketches as well as those unique to each sketch can be separately identified, thus allowing a mixing of different sketches to generate a topologically and geometrically rich set of conceptual alternatives. We demonstrate our technique with 2D and 3D examples.

Copyright © 2012 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Figure 1

Proposed method works from user-provided conceptual sketches. It then develops these sketches into a design space. The design space is either manually explored or used in integrated shape optimization.

Grahic Jump Location
Figure 2

Geometric representation of sketches. (a) Input sketch. (b) A subset of stroke groups, each defining an individual curve. (c) Vectorized curves. (d) T and L joints defined by the user. (e) Resulting line drawing.

Grahic Jump Location
Figure 3

(a) A line drawing of a flash drive. (b) Corresponding graph representation. M denotes that the curve-joint link corresponds to a midpoint connection on the curve, E denotes an end point connection. The graph has 9 curve nodes and 12 joint nodes.

Grahic Jump Location
Figure 4

(a) The sketch graphs are decomposed into base and detail subgraphs. (b) The base graphs are used to define a design space in order to synthesize new shapes from geometric variations. (c) The resulting shape is then recombined with previously decomposed details.

Grahic Jump Location
Figure 5

The basic geometric properties used for pairwise dissimilarity features. Each vectorized sketch is uniformly scaled into a unit square prior to feature calculation

Grahic Jump Location
Figure 6

A nonlinear scaling function that is used to combine curve scores in order to compare graph matches with different sizes

Grahic Jump Location
Figure 7

Curve directions are corrected using the corresponding joint information

Grahic Jump Location
Figure 8

Transformation between exemplars: (a) Each line segment on both curves is converted to a triangle. (b) jth triangle in C1 then undergoes an affine transformation to produce the corresponding triangle in C2 .

Grahic Jump Location
Figure 9

Shape synthesis using a nonlinear design space: (a) While deformation from the first exemplar to the second (b) can be interpolated with either (left) positive or (middle) negative amounts, a nonlinear design space can also handle (right) extrapolations involving large deformations

Grahic Jump Location
Figure 10

Sketch modification using automatic weight calculation propagates local modifications to the remainder of the sketch using the geometric deformations learned from the exemplars

Grahic Jump Location
Figure 11

Design of a new mug from four different designs

Grahic Jump Location
Figure 12

Design of a new car from four different designs

Grahic Jump Location
Figure 13

Design of a new hair dryer from three different designs

Grahic Jump Location
Figure 14

Synthesis from 3D wireframes. (Top row) Analysis of three handheld device designs represented as 3D curve networks for canonical form (dark) and feature (light) identification. The three designs and the selected features are combined to produce a new design. (Bottom row) Surfaced exemplars and the final design.

Grahic Jump Location
Figure 15

(a) and (b) Bottle design consisting of soda, wine, and beer bottles as exemplars. The bottle is formed by revolving the sketched contours. The size variation among the exemplars should be noted. (c) Example design obtained by exemplar weight control. (d) Geometric constraints. (e) Optimum design and resulting exemplar weights.

Grahic Jump Location
Figure 16

Mouse design from top views and side views. Internal and external constraints are imposed. Optimization aims to minimize interference with these constraints.

Grahic Jump Location
Figure 17

Minimizing curvature variation. A design which has the least variation of curvature is sought.

Grahic Jump Location
Figure 18

The vectorization of sketches directly impact the resulting graph structure. (a) Strokes that define continuous or disconnected sets (b) are converted into single or multiple curve segments (c) resulting in different graph structures. (d) Matching of geometrically similar, (e) yet topologically different curves is possible by simplifying graphs through forming compound curves.

Grahic Jump Location
Figure 19

Visual similarities (shown in red circles) among complicated geometries is difficult to detect with corresponding curves at the local level.

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In