Abstract
An evolutionary design method of image filter based on hardware-in-the-loop (HIL) simulation is proposed. An image filter evolvable system based on System Generator was designed. Experimental results and analyses demonstrate that the optimal evolved filter outperforms traditional filters both on performance and implementation costs. In comparison with the existing evolutionary design methods of image filter, the proposed strategy brings higher computational ability and more flexibility. Evolutionary algorithm of the evolvable system can be synthesized into hardware which provided a basis for the follow-up research to achieve online adaptive filtering. DOI: http://dx.doi.org/10.5755/j01.eee.20.7.3375
Highlights
Image filters widely used in the pre-processing phase, are an important part of the image processing systems
Traditional design process of image filter is mostly based on experimental work with the images and it leads to a very time consuming job [1]
Several studies are being focused on the evolutionary design of image filter [1], [8]–[12]. These methods can be classified into different classes based on the location of the fitness evaluation module and the evolutionary algorithm: 1. Extrinsic hardware evolution (EHE)—The image filters were evolved only by using a virtual reconfigurable circuit (VRC) simulated in software, no evolution in hardware was performed [1]
Summary
Image filters widely used in the pre-processing phase, are an important part of the image processing systems. Several studies are being focused on the evolutionary design of image filter [1], [8]–[12] These methods can be classified into different classes based on the location of the fitness evaluation module and the evolutionary algorithm: 1. 2. PC-based intrinsic hardware evolution (PCIHE)—The candidate circuits were evaluated in a reconfigurable device while EA was executed on a PC or a cluster of workstations [9], [10]. PC-based intrinsic hardware evolution (PCIHE)—The candidate circuits were evaluated in a reconfigurable device while EA was executed on a PC or a cluster of workstations [9], [10] We proposed a hardware-in-the-loop (HIL) simulation based evolutionary design method of image filter.
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call