Map of the world
Agriculture
Ecology and land management
Humanitarian interventions
Applied Meteorology
Aquatic environment
Regional and city planning
  Change detection
  Results
  Info
  On a town scale
  Land use
  Tourisme
Environmental risks
Global monitoring
Hyperspectral applications
Context

HOW TO IDENTIFY RECENT CHANGES IN THE URBAN LANDSCAPE?

 

The whole of Belgium on a scale of 1 : 10,000

An important task of the NGI (National Geographic Institute) is to digitally map the whole of Belgium on a scale of 1:10,000. This is a particularly interesting scale as it provides an overview of the whole territory while showing enough details at the same time. However, when dealing with large scale surveys, the question of updating becomes very important. At this level, the reality on the ground is constantly changing, especially the built-up areas and the road network, and to effectively meet the demand for geographical information from potential users this information must be updated regularly.

SPOT-5 image for the Sint-Niklaas region.
Change detection map for the Sint- Niklaas region.
Detecting changes

For detecting signifi cant recent changes in urbanisation, the NGI would like to have semi-automatic procedures that would present a considerable time-saving. To that end, the team from the Royal Military Academy set about testing the potential of satellite imagery. The NGI database, compiled in 2002, was compared with images acquired by the SPOT-5 satellite in 2004 and 2005, focusing on a number of small areas, some semi-urban and others rural. One of the challenges of this new approach lies in the maximal reduction of “false alarms”, that is the erroneous fl agging of change in urbanisation that quite simply proves to be wrong.

A 5m resolution

The potential of SPOT-5 images (2.5 or 5m spatial resolution) and of IKONOS images (1m resolution) was compared. The conclusion was that a 5m resolution was suffi cient to readily distinguish details of the road network and constructions of a certain size. As these are the urban features that change most quickly, they are also instrumental in steering management and planning decisions.

Objectives
To provide the NGI with an operational prototype for the automatic detection of significant changes in the road network and built-up areas, that require priority updating of the NGI database on a scale of 1:10,000.