3d Morphing And Shape Transformation Using Slices
Loading...
Date
2009-12
Authors
Yasmin, Shamima
Journal Title
Journal ISSN
Volume Title
Publisher
Universiti Sains Malaysia
Abstract
Generally the term ‘Shape Transformation’ means transformation of one object into another
object. When shape transformation involves two objects i.e. source object and destination object
and there is gradual, continuous and simultaneous dissolvement of the shape of source
object to destination object, it is called ‘Morphing’. Most of the existing shape transformation
algorithms involve a lot of user intervention, do not scale well when the number of input objects
is more than two and are not scalable in parallel and distributed computing environment
whenever needed. This thesis addresses the above-mentioned issues by proposing a novel 3D
morphing algorithm using slices. Firstly source and destination objects are traversed and slices
are generated from this traversal. Traversal of a particular data may generate slices which may
not be parallel to each other. Equal number of slices is generated for both source and destination
objects. Boundaries of both objects are extracted. The boundaries are then traversed
on a 2D plane and are transformed into a cloud of contour points each. Interpolation of the
corresponding source and destination contour point clouds takes place and transformed surface
is reconstructed.
The algorithm is dynamic and flexible enough as the number of slices and the number of
traversals can be varied. As Oriented Bounding Box (OBB) is used to determine the initial
alignment of the object, the algorithm automatically takes into account of the curvature/distortion
within an object to a certain extent. The algorithm involves the least amount of user
intervention. Instead of binding the concept of shape transformation only between ‘source’ and ‘destination’ objects, the method is also extended to the concept of shape transformation
involving more than two objects. Different objects can also be blended in different proportions
in the transformed output. As a spin-off of the basic work on shape transformation, a novel algorithm
on surface reconstruction from slices is also developed. It has also been shown that this
slice-based method scales well in parallel/distributed computing environment. Lastly, the ease
in which fully feature-based method can also be incorporated in the algorithm is shown. Complexity
of the algorithm is also analysed and the result shows better performance in comparison
with other shape transformation algorithms.
Description
Keywords
3d Morphing And Shape Transformation , Using Slices