Faculty of Information Technology, BUT

Publication Details

Self-Reconfigurable Evolvable Hardware System for Adaptive Image Processing

SALVADOR Ruben, OTERO Andres, MORA Javier, DE la Torre Eduardo, RIESGO Teresa and SEKANINA Lukáš. Self-Reconfigurable Evolvable Hardware System for Adaptive Image Processing. IEEE Transactions on Computers, vol. 62, no. 8, pp. 1481-1493. ISSN 0018-9340.
Czech title
Seberekonfigurovatelný evolvable hardware systém pro adaptivní zpracování obrazu
Type
journal article
Language
english
Authors
Salvador Ruben (UPN)
Otero Andres (UPN)
Mora Javier (UPN)
De la Torre Eduardo (UPN)
Riesgo Teresa (UPN)
Sekanina Lukáš, prof. Ing., Ph.D. (DCSY FIT BUT)
Keywords
evolutionary computing, genetic algorithms, evolvable hardware, FPGAs, self-adaptive systems, reconfigurable hardware, adaptable architectures, autonomous systems
Abstract
This paper presents an evolvable hardware system, fully contained in an FPGA, which is capable of autonomously generating digital processing circuits, implemented on an array of processing elements (PEs). Candidate circuits are generated by an embedded evolutionary algorithm and implemented by means of dynamic partial reconfiguration, enabling evaluation in the final hardware. The PE array follows a systolic approach, and PEs do not contain extra logic such as path multiplexers or unused logic, so array performance is high. Hardware evaluation in the target device and the fast reconfiguration engine used yield smaller reconfiguration than evaluation times. This means that the complete evaluation cycle is faster than software-based approaches and previous evolvable digital systems. The selected application is digital image filtering and edge detection. The evolved filters yield better quality than classic linear and nonlinear filters using mean absolute error as standard comparison metric. Results do not only show better circuit adaptation to different noise types and intensities, but also a nondegrading filtering behavior. This means they may be run iteratively to enhance filtering quality. These properties are even kept for high noise levels (40 percent). The system as a whole is a step toward fully autonomous, adaptive systems.
Published
2013
Pages
1481-1493
Journal
IEEE Transactions on Computers, vol. 62, no. 8, ISSN 0018-9340
Publisher
IEEE Computer Society
DOI
BibTeX
@ARTICLE{FITPUB10105,
   author = "Ruben Salvador and Andres Otero and Javier Mora and Eduardo Torre la De and Teresa Riesgo and Luk\'{a}\v{s} Sekanina",
   title = "Self-Reconfigurable Evolvable Hardware System for Adaptive Image Processing",
   pages = "1481--1493",
   journal = "IEEE Transactions on Computers",
   volume = 62,
   number = 8,
   year = 2013,
   ISSN = "0018-9340",
   doi = "10.1109/TC.2013.78",
   language = "english",
   url = "https://www.fit.vut.cz/research/publication/10105"
}
Back to top