Image Analysis Used to Measure the Spatial Complexity of Land Cover Across Coastal Zone
Christine Voiron-Canicio,
UMR 6012 ESPACE - CNRS / Université de Nice Sophia-Antipolis (FR)
During the last fifty years the increasing attractivity
of Mediterranean regions has brought about important land use changes so that
the organization and differentiation of coastal spaces have became more and
more complex.
In this paper we chose to focus on spatial configurations of coastal land
uses, usig CORINE land cover images and taking into account both spatial and
temporal dimensions. More precisely, the aim of this contribution is to present
a method for detecting and measuring land cover spatial complexity.
The spatial complexity is relevant to the way land use classes are interlaced
and it is defined as a high degree of land use categories interweaving.
In a first stage, we discusse the interest and the limits of Gis-tools in
detecting spatial configurations. Those tools belong to counting statistics
family and they do not take into account the way geographical objects are
interlaced to each other across the space. Next, spatial and morphological
tools related to image analysis are presented. and Mathematical Morphology
operators are compared with Gis-tools.
In a second stage, a method for measuring the spatal complexity of coastal
land cover is proposed and applied to CORINE land cover and LACOAST images
of Côte d'Azur and Liguria. The approach consists in :
- converting CORINE and LACOAST images into grey-tones images,
- analyzing how land use components are interwoven, using image transformations which provide different shape measures, as fractal measure, and spatial complexity indicators,
- comparing spatial complexity results of the two coastal regions, at different dates,
- examining the observed changes in order to search their significance and to understand the processes at work.