Homepage | Links | Contacts
Projects

Duration: 12 months)
Starting date: May 2014
Project Contact:
Antonio Falcão
(ajf@uninova.pt)

Technical officer:
Peter van der Plas

Financing:
This project is financed by ESA under the ITI (Triangle Initiative) programme. ESA contract 4000111155/14/NL/SC

IVELA - Interactive Visualization Environment for Large Archives
Project Summary
In an age when automated massive data analysis is becoming a mainstream activity, the task of visualization is essential for the understanding and critical interpretation of data and results - for science and mission analysis and control alike. Particularly, in Space related activities, most of the data leading to new scientific discoveries are expected to come from online archives of automatic instruments and all-sky surveys.

The main objective of IVELA is to develop technologies to tackle two main challenges: how to interactively visualize a large amount of data and how to perform complex selections in 3D environments.

The emergence of widespread 3D interaction devices such as Microsoft Kinect and Leap Motion, brought a hope that perhaps this problem can be soon addressed.

Those devices use machine learning algorithms with image processing and fusion, for enabling the computer to recognize humans and to perform motion capture.

However, for effective usage in data selection, the system should also be able to infer the user's motion intention. It should be able to answer questions as: does the user want to select a sphere, or an arbitrarily complex region? Hence, IVELA will recognize gestures that humans intuitively use for data selection and volume definition. This will involve the adoption of machine learning algorithms (as Support Vector Machines), to infer the regions from the user's gesture.

Currently, there is no scientific, engineering or mission analysis software tool that allows the user to perform data selection of complex regions in the 3D space using such human-computer interaction devices.

Finally, even if IVELA is a demonstrator for a 3D interactive visualization environment for large datasets such as Gaia and space debris datasets, its design will be as general as possible, and it should be easily adapted for the visualization of other space data sources.

Video demo #1 - functionalities of the visualisation client tool



Video demo #2 - interaction using the Leap Motion device

The project tasks are:

WP1: preliminary studies
WP2: visualization demonstrator infrastructure
WP3: advanced data selection in 3D
WP4: system integration
WP5: benchmarking
WP6: management

Specifically, UNINOVA will participate in all tasks and will be responsible for WP3.

Research Topics
  • Data mining
  • Visual analytics
  • Big data visualization
Partners