[ Les effets du niveau de l'échelle et de la résolution spatiale sur la classification orientée objet : Application à la cartographie de l'Arganeraie (Région Agadir, Maroc) ]
Volume 14, Issue 1, March 2015, Pages 21–31
O. El Kharki1, J. Mechbouh2, M. Rouchdi3, A. EL ABOUDI4, and J. M. NGONO5
1 Cellule informatique, Université Ibn Zohr, ENCG, Agadir, Maroc
2 Centre Régional Africain des Sciences et Technologies de l'Espace, Rabat, Maroc
3 IAV Hassan II, Rabat, Maroc
4 Faculté des Sciences Rabat, Maroc
5 Académie Internet, Université de Douala, Cameroun
Original language: French
Copyright © 2015 ISSR Journals. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The Argan tree in southwest Morocco is characterized by fragmentation of its forest area and by its adaptability to all types of soils, which disrupts its spectral response. Owing to the complexity in these areas, mapping of the Argan tree facing serious problems when applying methods based on spectral information and ignore spatial information. In this paper, we examine the performance of the object -oriented approach with two satellite images of different spatial resolution (high and medium resolution). Next, we compare the results of object-oriented classification with those produced by the pixel approach. Finally, we evaluate the performance of the object-oriented classification with respect to different segmentation levels. This study shows that object oriented classification is more performance compared to the pixel approach for high spatial resolution images. The results obtained proved also that the object-oriented classification depended on the segmentation scale.
Author Keywords: spatial Information, spectral information, classification, object-oriented, pixel approach, scale level, spatial resolution.
Volume 14, Issue 1, March 2015, Pages 21–31
O. El Kharki1, J. Mechbouh2, M. Rouchdi3, A. EL ABOUDI4, and J. M. NGONO5
1 Cellule informatique, Université Ibn Zohr, ENCG, Agadir, Maroc
2 Centre Régional Africain des Sciences et Technologies de l'Espace, Rabat, Maroc
3 IAV Hassan II, Rabat, Maroc
4 Faculté des Sciences Rabat, Maroc
5 Académie Internet, Université de Douala, Cameroun
Original language: French
Copyright © 2015 ISSR Journals. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
The Argan tree in southwest Morocco is characterized by fragmentation of its forest area and by its adaptability to all types of soils, which disrupts its spectral response. Owing to the complexity in these areas, mapping of the Argan tree facing serious problems when applying methods based on spectral information and ignore spatial information. In this paper, we examine the performance of the object -oriented approach with two satellite images of different spatial resolution (high and medium resolution). Next, we compare the results of object-oriented classification with those produced by the pixel approach. Finally, we evaluate the performance of the object-oriented classification with respect to different segmentation levels. This study shows that object oriented classification is more performance compared to the pixel approach for high spatial resolution images. The results obtained proved also that the object-oriented classification depended on the segmentation scale.
Author Keywords: spatial Information, spectral information, classification, object-oriented, pixel approach, scale level, spatial resolution.
Abstract: (french)
L'arganier du sud-ouest Marocain, est caractérisé par la fragmentation de son espace forestier ainsi que par son adaptabilité à tous les types des sols, ce qui perturbe sa réponse spectrale. En raison de la complexité dans ces zones, la cartographie de l'arganeraie rencontre de sérieux problèmes lors de l'application des méthodes basées sur l'information spectrale et qui ignorent l'information spatiale. Dans ce papier, nous examinons les performances de l'approche orientée objet avec deux images satellites de résolution spatiale différente (haute et moyenne résolution). Ensuite, Nous comparons les résultats de la classification orientée objet avec ceux produits par l'approche pixel. Enfin, nous évaluons les performances de la classification orientée objets par rapport à différentes échelles de segmentation. Cette étude montrent que la classification orientée objet n'est plus performante par rapport à l'approche pixel que pour les images de haute résolution spatiale. Les résultats obtenus mettent en évidence la dépendance de la classification orientée objet avec l'échelle de segmentation.
Author Keywords: Information spatiale, information spectrale, classification, orientée objet, approche pixel, échelle de segmentation, résolution spatiale.
How to Cite this Article
O. El Kharki, J. Mechbouh, M. Rouchdi, A. EL ABOUDI, and J. M. NGONO, “The effects of the scale level and of the spatial resolution on the object oriented classification : Application to mapping Arganeraie (region Agadir, Morocco),” International Journal of Innovation and Scientific Research, vol. 14, no. 1, pp. 21–31, March 2015.
