Podpora paralelizace pro makro jazyk ImageJ

Abstract

Fiji is an open-source platfrom based on ImageJ software that focuses on scientific image processing. Fiji lacks paralelization support and therefore the goal of this thesis is to develop and implement plugin that uses parallelization to apply convolution algorithm on image by using OpenMPI. Convolution consists of applying convolution filter on image and returns processed image. Detailed process of implementation is described and showcased. The plugin can be also used in ImageJ macro language. The implementation is then tested on supercomputer with variable input parameters.