The objective of the paper is to carry out the spatial distribution of the built-up areas in the Cameroonian shores of the Lake Chad and its hinterland through orientated-object classification. The methods used combines spectral and statistical processing’s. For the Principal Component Analysis (PCA) performed, the first component contains 87% of information. The built up areas indices have also computed, like Urban Index (UI) and Normalized Difference Built-up Index (NDBI) based on sentinel 2 satellite images of April 2017. The components 3 and 6 of PCA and the built up indices are stacked with originals spectral bands to generate a new composite image. This new image is then segmented and classified by Support Vector Machines (SVMs) algorithm, submitted to a threshold analysis. The result is a classified image on which the identification of built-up areas from the others spatial objects is possible and the spatial distribution of built-up areas is countable.
Volume 11 | 11-Special Issue