The datasets released for the VAST Challenge 2017 comprise vehicle movement data captured with RFID sensors, chemical emission data from factories captured by gas sensors, and image attributes of the wildlife plant health obtained from satellites, all pertaining to a fictional wildlife preserve. Using visual analytics, a compelling hypothesis is established to link the spatiotemporal datasets to the phenomenon, where the count of a bird specimen is found to decline over a given year. Anomalies in vehicle traffic patterns are linked to proximal factory emissions, and further associated with satellite imagery that show proof of degradation in plant quality in the preserve. The evidences are supported with visualizations created in Tableau, R, QGIS & SAS-JMP. Raster image analysis is also done to identify other key features in the preserve, such as the existence of a lake. This is achieved by using NDVI and NDMI measures, which also help understand the change in climate over the years.
Spatiotemporal analysis, Geo-spatial analytics, Visual analytics, Traffic pattern detection, Raster image processing, NDVI
Databases and Information Systems | Geographic Information Sciences
Data Management and Analytics
IEEE VAST Challenge, Phoenix, Arizona, 2017 October 1-6
Singapore Management University
City or Country
KISHAN, Bharadwaj; ONG, Jason Guan Jie; ZHANG, Yanrong; and KAM, Tin Seong.
Spatiotemporal identification of anomalies in a wildlife preserve. (2017). IEEE VAST Challenge, Phoenix, Arizona, 2017 October 1-6. Research Collection School Of Information Systems.
Available at: http://ink.library.smu.edu.sg/sis_research/3831
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.