• TUCS Publication Series
• How to Publish in TUCS Series
• Publication Search
• Publication Graph
• Publication Input
• Publication Input Guide
• JuFo Browser

Developing and testing structural light vision software by co-evolutionary genetic algorithm

Timo Mantere, Jarmo T. Alander, Developing and testing structural light vision software by co-evolutionary genetic algorithm. In: QSSE 2002 The Proceedings of the Second ASERC Workshop on Quantative and Soft Computing based Software Engineering, 31-37, Alberta Software Engineering Research Consortium (ASERC) and the Department of Electrical and Computer Engineering, University of Alberta, 2002.

Abstract:

In this paper we propose an approach to automatically develop and
test software by co-evolutionary optimization using genetic algorithms. The
idea is to generate both rule based methods for combining scan image data
and corresponding simulated test surfaces for a structured light volume
measurement system. The goal is to minimize the worst case behavior of
error bounds of the volume measurement. One genetic algorithm is used to
generate rules to combine scan data that give minimum relative error on the
test surface population, which is generated by another genetic algorithm
trying to create surfaces giving high measurement error. Thus the surface
population defines the fitness of the method population and vice versa.
Based on observations of evolution in nature it is believed that it is this kind
of co-evolution that leads in the long run to excellent solutions, that would be
difficult to find by more traditional genetic algorithm approaches. Indeed, the
preliminary results got seem to indicate that co-evolution is beneficial in
software development and testing.

BibTeX entry:

@INPROCEEDINGS{pMaAl02a,
  title = {Developing and testing structural light vision software by co-evolutionary genetic algorithm},
  booktitle = {QSSE 2002 The Proceedings of the Second ASERC Workshop on Quantative and Soft Computing based Software Engineering},
  author = {Mantere, Timo and Alander, Jarmo T.},
  publisher = {Alberta Software Engineering Research Consortium (ASERC) and the Department of Electrical and Computer Engineering, University of Alberta},
  pages = {31-37},
  year = {2002},
  keywords = {Co-evolution, genetic algorithms, image processing, 3-D imaging, machine vision, simulation, software engineering, software testing, structured light},
}

Edit publication