Rosalina Abdul Salam
Universiti Sains Malaysia
School of Computer Science
11800 Penang
Malaysia
|
e-mail: Rosalina@cs.usm.my |
|
Marcos Aurelio Rodrigues
University of Sheffield Hallam
School of Computing and Management Sciences
S1 1WB Sheffield
United Kingdom
e-mail: m.rodrigues@shu.ac.uk
Keywords: . Shape outline, viewpoint dependent, multiple viewpoints, classification, recognition, three-dimensional.
Abstract
The
design of a general purpose artificial vision system capable of recognizing
arbitrarily complex three-dimensional objects without human intervention is
still a challenging task in computer vision. Computer vision research has tried
to incorporate knowledge of how human vision works and use this knowledge to
design robust recognition systems. Early vision system, that is the primary
visual cortex is where the edge and bar detection happen. These knowledge on
how human vision works can be use to design a robust recognition system.
Experiments have been conducted by incorporating these knowledge. Firstly, the
process of shape outline detection and secondly, the use of multiple viewpoints
of object. Shape outline readings are put through a normalization and
dimensionality reduction process using an eigenvector based method to produce a
new set of readings. Through statistical
analysis, these readings together with
other key measures, namely peak measures
and distance measures, a robust
classification and recognition process is achieved. Tests show that the
suggested methods are able to automatically recognize three-dimensional objects
from multiple viewpoints. Finally, experiments also demonstrate the system
invariance to rotation, translation, scale, reflection and to a small degree of
distortion. Tests also show that the suggested methods are able to
automatically recognize three-dimensional objects from multiple viewpoints
without any extra information required during the whole process.