DocumentCode
2746705
Title
A multi-data sources method for the validation of old image classification in order to measure the scattered urbanisation process
Author
Maille, E.
fYear
2003
fDate
22-23 May 2003
Firstpage
152
Lastpage
156
Abstract
This article proposed a mapping and validation method of old low resolution satellite images, based on the use of existing GI covers at different scales for different periods. This method can be described as a multi data sources method for the validation of old image classification in order to measure the scattered urbanisation process as it uses three kinds of data: satellite images, GI vectorial covers, and digitised paper maps. The hypothesis is that the satellite image allows to map the exact borders of homogeneous patches, while small scale vector GI cover provides the exact content. That is why satellite images are used to increase the precision of the GI cover. However, the content of homogeneous patches detected on the images is not totally unknown, as absolute radiometric values and visual texture analysis on coloured band combination gives some information.
Keywords
image classification; image colour analysis; image resolution; image texture; signal detection; terrain mapping; GI vectorial covers; absolute radiometric values; coloured band combination; digitised paper maps; homogeneous patches borders; image detection; low resolution satellite images; multidata sources method; old image classification validation; scattered urbanisation process; visual texture analysis;
fLanguage
English
Publisher
ieee
Conference_Titel
Remote Sensing and Data Fusion over Urban Areas, 2003. 2nd GRSS/ISPRS Joint Workshop on
Conference_Location
Berlin, Germany
Print_ISBN
0-7803-7719-2
Type
conf
DOI
10.1109/DFUA.2003.1219977
Filename
5731019
Link To Document