Remote sensing: Digital image processing (3/3)


After the acquisition of satellite images, these must undergo processing: 

Radiometric correction: it consists in correcting the errors produced by sensor imperfections, the effect of the atmosphere ... The correction of these errors is based on mathematical models, or else an algebraic correction (we subtract for example a minimum radiometric value of all pixels). 

Georeferentiation: this is a geometric transformation. This process aims to coincide the image coordinate system with a reference system. This corrects the image of the effect of the relief and the tilt of the aerial platform. So that it can be superimposed on a topographic map. Like any geometric transformation, a resampling is necessary, as the extent of the pixels is modified and therefore a radiometric value must be associated with the new pixels, the choice of the resampling technique depends on the purpose of the study: 

- Nearest neighbor: allows you to associate with each new pixel the value of the old pixel which is closest to it. 

- Bilinear transformation: interpolation of the values of the 4 neighboring pixels. 

- Cubic transform: double interpolation, it uses the 16 pixels neighboring the pixel concerned. 

Enhancement: it is a radiometric transformation. This operation consists in making the image more contrasted, the histogram (represents the number of occurrences of the radiometric values) then becomes more distributed and uniform. There are several varieties of enhancement, in the general case, a linear enhancement is applied. However, it is possible to apply an enhancement in stages serving to apply an enhancement coefficient to each range of radiometric values. 

Filtering: consists in applying a filter to the image to highlight particular details, we often hear two types of filters: 

- High pass filter: allows to highlight areas with high spatial frequency (contours, linear details ...) 

- Low pass filter: allows the highlighting of zones with low spatial frequency (uniform zones ...)

Thresholding: it is an operation which consists in fixing thresholds for certain phenomena. To fully understand, we suppose that we want to know the plant areas from the non-plant areas, for that, we must apply a threshold which makes it possible to distinguish the plant areas. For example, we apply a threshold of 200 for the reflectance in green. This means that the zones whose reflectance of green exceeds 200 are zones deemed to be vegetation, and the others are not vegetation. (There is an index that most characterizes vegetation called NDVI: normalized difference vegetation index) 

Classification: it is an operation which is used to extract the classes in the image based on the radiometric values. That is to say extracting areas of vegetation, water (lakes, sea ...), urban, bare ground. For quality control, we go out to the field to check the class of a sample of points distributed over the study area. Classification can be supervised (requires prior knowledge of the field) or unsupervised (100% automatic).

After classification, we get a land use map.





Commentaires

Enregistrer un commentaire

Posts les plus consultés de ce blog

L'urbanisme au Maroc et les documents d'urbanisme

Image
Urbanisme : C’est l’ensemble des techniques et sciences permettant d’aménager les espaces urbains et de tracer les grandes lignes et les orientations de l’urbanisation. Le Maroc s’est très tôt doté de plusieurs instruments afin d’organiser l’urbanisation des villes et cadrer la planification urbaine de ses dernières. Ceux-ci se matérialisent en documents d’urbanisme dont l’objectif est l’appréhension des modes d’utilisation des sols, et plus généralement de répondre spatialement aux objectifs tracés en ce qui concerne les grandes questions du logement, infrastructure équipement… Lois régissant l’urbanisme au Maroc :      Le Dahir du 16 avril 1914      Le Dahir du 12 novembre 1917 relatif aux associations syndicales de propriétaires urbains      Loi du 27 janvier 1931 autorisant l’établissement de Plans d’aménagement pour les centres et les banlieues des villes      Dahir du 14 juin 1933 relatif aux lotissement ...
Lire la suite

Les points géodésiques: définition, utilités et difficultés ...

Image
     Ceux qui ont visité Toubkal ont surement remarqué un  quadripode  la dessus, certains même ont essayé de l’escalader d’enthousiasme. Et peu sont ceux qui se demandent de quoi s’agit t-il? C’est tout simplement un  point géodésique !     Un point géodésique est un point sur la surface ou sur des  monopodes -  bipodes -  tripodes  ou  quadripodes  dont les coordonnées sont connues avec  précision . Ces points ont une grande importance topographique vu que: - Ils constituent la base des  levés  top ographiques. - Ils sont la base de  densification  d’une zone déterminée. - Ils sont importants au voisinage des  ouvrages d’arts  qui peuvent être servis soit pour effectuer des levés, des implantations, ou un contrôle de stabilité. - Peuvent servir de base pour des travaux  GNSS  ou  LiDaR . - Utilisation  photogrammétrique   comme GCP ( Ground Contro...
Lire la suite

Remote sensing: basics, sensor’s characteristics, image composition ... (2/3)

Image
We have to notice that the radiation sent, before arrival at the object on the surface, it is subjected to an interaction with the atmosphere (which can reflect a part) and even after the reflection on the object it enters into a second interaction with the atmosphere, and after that, the radiation is captured by sensors. To remedy this problem of interaction with the atmosphere, we exploit the characteristics of electromagnetic radiation, for example we choose wavelengths that allow the penetration of the atmosphere without any reflection on it (atmospheric window) After scanning and recording the reflectance values, these values are sampled to have gray levels from 0 (black) to 255 (white). The basic unit is the pixel, each pixel has a radiometric value for each band (value for green, red, blue, infrared, ...). The final image is built in the form of a pixel matrix. In which every pixel has a radiometric value. A satellite sensor has 4 fundamental characteristics:  Spatial resolu...
Lire la suite