Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Evolutionary optimisation of a morphological image processor for embedded systems
University of Borås, School of Engineering.
2008 (English)Doctoral thesis, monograph (Other academic)
Abstract [en]

The work presented in this thesis concerns the design, development and implementation of two digital components to be used, primarily, in autonomously operating embedded systems, such as mobile robots. The first component is an image coprocessor, for high-speed morphological image processing, and the second is a hardware-based genetic algorithm coprocessor, which provides evolutionary computation functionality for embedded applications. The morphological image coprocessor, the Clutter-II, has been optimised for efficiency of implementation, processing speed and system integration. The architecture employs a compact hardware structure for its implementation of the morphological neighbourhood transformations. The compact structure realises a significantly reduced hardware resource cost. The resources saved by the compact structure can be used to increase parallelism in image processing operations, thereby improving processing speed in a similarly significant manner. The design of the Clutter-II as a coprocessor enables easy-to-use and efficient access to its image processing capabilities from the host system processor and application software. High-speed input-output interfaces, with separated instruction and data buses, provide effective communication with system components external to the Clutter-II. A substantial part of the work presented in this thesis concerns the practical implementation of morphological filters for the Clutter-II, using the compact transformation structure. To derive efficient filter implementations, a genetic algorithm has been developed. The algorithm optimises the filter implementation by minimising the number of operations required for a particular filter. The experience gained from the work on the genetic algorithm inspired the development of the second component, the HERPUC. HERPUC is a hardware-based genetic algorithm processor, which employs a novel hardware implementation of the selection mechanism of the algorithm. This, in combination with a flexible form of recombination operator, has made the HERPUC an efficient hardware implementation of a genetic algorithm. Results indicate that the HERPUC is able to solve the set of test problems, to which it has been applied, using fewer fitness evaluations and a smaller population size, than previous hardware-based genetic algorithm implementations.

Place, publisher, year, edition, pages
Chalmers University of Technology, Dep. of Applied Mechanics , 2008.
Series
Doktorsavhandlingar vid Chalmers tekniska högskola. Ny serie, ISSN 0346-718X ; 2759
Keywords [en]
embedded system, machine vision, morphological image processing, genetic algorithm
Keywords [sv]
Energi och material
National Category
Computer Sciences Computer Vision and Robotics (Autonomous Systems) Other Electrical Engineering, Electronic Engineering, Information Engineering
Identifiers
URN: urn:nbn:se:hb:diva-3479Local ID: 2320/3649ISBN: 978-91-7385-078-0 (print)OAI: oai:DiVA.org:hb-3479DiVA, id: diva2:876868
Available from: 2015-12-04 Created: 2015-12-04 Last updated: 2018-01-10

Open Access in DiVA

No full text in DiVA

Authority records BETA

Magnusson, Andreas

Search in DiVA

By author/editor
Magnusson, Andreas
By organisation
School of Engineering
Computer SciencesComputer Vision and Robotics (Autonomous Systems)Other Electrical Engineering, Electronic Engineering, Information Engineering

Search outside of DiVA

GoogleGoogle Scholar

isbn
urn-nbn

Altmetric score

isbn
urn-nbn
Total: 127 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf