Parallel image sharpness measure for supercomputing environment

Abstract

Image sharpness is an important part of the overall quality of an image. There are multiple metrics to evaluate image sharpness, the algorithms are mainly created for regular desktop computers. This article describes the sharpness metrics devised and implemented at Obuda University parallelized and ported to an HPC infrastructure. Along with the native ported application a workflow has been created on the gUSE/WS-PGRADE portal system and an ASM portlet has been developed to make the service readily available to I.