Pavel Smrz, BUT
Ondrej Klima, BUT
This experiment aims at finding an optimal distribution of the reconstruction algorithm across the cluster of computation nodes. The experimental application will be executed many times with different numbers of nodes and processes running on each computation node. Time necessary for each step of the algorithm – the extraction of SIFT features, matching these features between corresponding stereo images and tracking them in neighbour images – will be preserved for each input picture in the XML metadata and then processed by Hadoop-based tools. Results of the metadata analysis will characterize the average computation time of each mentioned processing stage.
Policy statements that relate to this experiment and any evaluation criteria taken from SCAPE metrics
Links to results of the experiment using the evaluation template.