Scientific Visualization of Radio Astronomy Data using Gesture Interaction.

Mulumba, Patrick and Gain, James and Marais, Patrick and Woudt, Patrick (2015) Scientific Visualization of Radio Astronomy Data using Gesture Interaction., Proceedings of Astronomical Data Analysis Software and Systems XXIV (ADASS XXIV), September 2015, 145-145.

Full text not available from this repository. (Use alternate locations listed below)

Abstract

MeerKAT in South Africa (Meer = More Karoo Array Telescope) will require software to help visualize, interpret and interact with multidimensional data. While visualization of multi-dimensional data is a well explored topic, little work has been published on the design of intuitive interfaces to such systems. More specifically, the use of non-traditional interfaces (such as motion tracking and multi-touch) has not been widely investigated within the context of visualizing astronomy data. We hypothesize that a natural user interface would allow for easier data exploration which would in turn lead to certain kinds of visualizations (volumetric, multidimensional). To this end, we have developed a multi-platform scientific visualization system for FITS spectral data cubes using VTK (Visualization Toolkit) and a natural user interface to explore the interaction between a gesture input device and multidimensional data space. Our system supports visual transformations (translation, rotation and scaling) as well as sub-volume extraction and arbitrary slicing of 3D volumetric data. These tasks were implemented across three prototypes aimed at exploring different interaction strategies: standard (mouse/keyboard) interaction, volumetric gesture tracking (Leap Motion controller) and multi-touch interaction (multi-touch monitor). A Heuristic Evaluation revealed that the volumetric gesture tracking prototype shows great promise for interfacing with the depth component (z-axis) of 3D volumetric space across multiple transformations. However, this is limited by users needing to remember the required gestures. In comparison, the touch-based gesture navigation is typically more familiar to users as these gestures were engineered from standard multi-touch actions. Future work will address a complete usability test to evaluate and compare the different interaction modalities against the different visualization tasks.

Item Type: Conference paper
Date Deposited: 13 Sep 2016
Last Modified: 10 Oct 2019 15:32
URI: http://pubs.cs.uct.ac.za/id/eprint/1096

Actions (login required)

View Item View Item