Delimiting imprecise regions with georeferenced photos and land coverage data

Authors:
Bruno Martins
Affiliations:
Instituto Superior Técnico INESC-ID, Porto Salvo, Portugal
Venue:
W2GIS'11 Proceedings of the 10th international conference on Web and wireless geographical information systems
Year:
2011

Citing 7
Cited 0

Estimating probability surfaces for geographical point data: an adaptive kernel algorithm

Computers & Geosciences
Estimation of High-Density Regions Using One-Class Neighbor Machines

IEEE Transactions on Pattern Analysis and Machine Intelligence
Modelling vague places with knowledge from the Web

International Journal of Geographical Information Science - Digital Gazetteer Research
Geographical information retrieval

International Journal of Geographical Information Science
Mapping the world's photos

Proceedings of the 18th international conference on World wide web
Multivariate outlier detection in exploration geochemistry

Computers & Geosciences
Images and perceptions of neighbourhood extents

Proceedings of the 6th Workshop on Geographic Information Retrieval

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents an automated method for defining the boundaries of imprecise regions with basis on publicly available data. The method combines interpolation from a set of points which are assumed to lie in the region to be delineated, obtained from Flickr and evaluated through Kernel Density Estimation, with heuristics for refining the results that leverage on land coverage datsets obtained through remote sensing, integrated through an approached based on region shrinkage. The overall approach is evaluated by means of statistical classification measures, using regions whose boundaries are well defined. Our results shows that the method proposed here performs better than previous approaches described in the litterature, based solely on interpolation through Kernel Density Estimation.