Résumé
183
Tab.1 List of data used in this study
Date
Data type
Image
scale
File type
Image type
1957 - 1959
Aerial photographs
1/25 000
Geo-TIFF
Black and
white
1973
Aerial photographs
1/25 000
Geo-TIFF
Black and
white
1980
Aerial photographs
1/10 000
Geo-TIFF
Black and
white
2006 - 2011 -
2016
Satellite image
1 - 10m
Geo-TIFF
Color
2012 - 2015 -
2017
DGPS topographic surveys
/
XYZ
/
2.4.Choice of the reference shoreline
The analysis of coastline mobility at the local level is based on several markers chosen
according to their relevance to the types, nature of available data, and sites studied. On the
aerial photographs, the upper limit of the range was selected to measure the evolution of the
coastline of the sites. On the other hand, in terms of satellite images, the only line used is the
high sea line before the aerial shots in 2006, 2011, and 2016.
2.5.Image processing and rectification methods
- External rectification
The digitized images contain various deformations inseparable from the mechanism of taking
the aircraft and the relief overflown. It is necessary to correct them to compare the extracted
coastlines to each other. Planimetric rectification (2D) corrects the radial distortions due to the
optics and those caused by the movements of the aircraft (pitch, roll, rollover) (Dolan et al.
1980 , Anders et al. 1990, Crowell et al. 1991, Moore, 2000). It involves applying a
rectification model by capturing homologous (bitter) points between two images, one being
the image to be corrected, the other is an already georeferenced image. Thereference is the
orthophoto-plan (1984) and the Topographic Map at 1:25,000, produced by INCT, with a
spatial resolution of 0.5 m and projected in UTM (WGS 1984 ellipsoid). The
orthophotographs result from a 3D rectification that corrects the vertical deformations related
to the relief, significant in continental areas, moderate on the coast (Dolan et al. 1978; Anders
et al. 1990, Crowell et al. 1991).
- Internal Rectifications (Image Correction)
The representation of an image on the screen is important for a good photo interpretation,
hence the importance of radiometric enhancement of an image in improving its visual
appearance. Our improvement was based on the examination of the histogram of each image
to mitigate the undesirable effects. But first, we had to assign a color to each of the three
channels (red, green, blue) for the multispectral satellite images. The command " band
combination " of the module " Raster " of Erdas Imagine was the tool of the attribution of the
colors to the channels (Red to 1, Green to 2, Blue to 3).
As for the histogram, a selection was made on a band of a few hundred meters to apply a
"general contrast meters to apply a "general contract" based on the "standard deviation of the
183
Tab.1 List of data used in this study
Date
Data type
Image
scale
File type
Image type
1957 - 1959
Aerial photographs
1/25 000
Geo-TIFF
Black and
white
1973
Aerial photographs
1/25 000
Geo-TIFF
Black and
white
1980
Aerial photographs
1/10 000
Geo-TIFF
Black and
white
2006 - 2011 -
2016
Satellite image
1 - 10m
Geo-TIFF
Color
2012 - 2015 -
2017
DGPS topographic surveys
/
XYZ
/
2.4.Choice of the reference shoreline
The analysis of coastline mobility at the local level is based on several markers chosen
according to their relevance to the types, nature of available data, and sites studied. On the
aerial photographs, the upper limit of the range was selected to measure the evolution of the
coastline of the sites. On the other hand, in terms of satellite images, the only line used is the
high sea line before the aerial shots in 2006, 2011, and 2016.
2.5.Image processing and rectification methods
- External rectification
The digitized images contain various deformations inseparable from the mechanism of taking
the aircraft and the relief overflown. It is necessary to correct them to compare the extracted
coastlines to each other. Planimetric rectification (2D) corrects the radial distortions due to the
optics and those caused by the movements of the aircraft (pitch, roll, rollover) (Dolan et al.
1980 , Anders et al. 1990, Crowell et al. 1991, Moore, 2000). It involves applying a
rectification model by capturing homologous (bitter) points between two images, one being
the image to be corrected, the other is an already georeferenced image. Thereference is the
orthophoto-plan (1984) and the Topographic Map at 1:25,000, produced by INCT, with a
spatial resolution of 0.5 m and projected in UTM (WGS 1984 ellipsoid). The
orthophotographs result from a 3D rectification that corrects the vertical deformations related
to the relief, significant in continental areas, moderate on the coast (Dolan et al. 1978; Anders
et al. 1990, Crowell et al. 1991).
- Internal Rectifications (Image Correction)
The representation of an image on the screen is important for a good photo interpretation,
hence the importance of radiometric enhancement of an image in improving its visual
appearance. Our improvement was based on the examination of the histogram of each image
to mitigate the undesirable effects. But first, we had to assign a color to each of the three
channels (red, green, blue) for the multispectral satellite images. The command " band
combination " of the module " Raster " of Erdas Imagine was the tool of the attribution of the
colors to the channels (Red to 1, Green to 2, Blue to 3).
As for the histogram, a selection was made on a band of a few hundred meters to apply a
"general contrast meters to apply a "general contract" based on the "standard deviation of the