A Method for Processing Satellite Image to Limit the Spread of Slums in the Regions of Marrakech

The slums are now a reality to be recognised. They were undoubtedly the prior nightmare of inhabitants in the world. Marrakech region experienced strong population growth. There is also a major investment in the tourism sector in the city and its outskirts. Unable to access any property, for lack of means, the newcomers live in these slums, and build shelters. What makes things more difficult? The solutions exist and are even multiple. First, shanty towns should be included in urban planning to allow for consistent management and better integration of informal settlements in the city. Then, plots could be created for these people to avoid disorganized construction. In this work, we are interested in detecting the boundaries of the villages near Marrakech from their satellite images, as a study case: Sidi Zouine village (31.67,8.34). Our approach has two stages: The first, using k nearest neighbour method for classification of the satellite image. The second, using a mean filter to regularize this classification. Our method gives satisfactory and encouraging results to detect the edge of the village and can be used as a tool to control the spread of slums.