Posters

Poster Session 2, Tuesday,  October  4, 10:40–12:40

Poster 24

Modification of WiPE (WiPE-2) for Land use/Land cover mapping based on combination of statistical of spectral and unsupervised classification: An application to VNREDSat-1 satellite image

Land use/Land cover (LULC) map is a necessary product for intensive thematic studies such as water, forest, land cover change, etc. WiPE was developed to extract only natural water pixels that a lot of contaminated water pixels with land pixels are masked. The identification of contaminated water pixels is essential to propose restoration solutions for improving the water surface information from the satellite image. Spectral shape analysis performs well in WiPE algorithm but it difficult to separate contaminated water pixels (sunglint, whitecaps, etc.) with other objects such as thin cloud, sand etc. To avoid the difficulty in classifying each pixel when the spectral amplitude of each noise object region on each scene is quite wide, the classification technique will help group these regions to form a base feature in the form of regions. Unsupervised classification, K-Mean, is effective when grouping pixels in areas of water, vegetation, and bare land, but it is difficult to create pixel regions in residential areas. Quickshift segmentation is a technique of clustering pixels by defining the kernel size and spectral difference in that block and this is the solution for grouping pixels in a residential area. Zonal statistic (min, max, mean, median, standard deviation) for each Rayleigh corrected spectral band at each classified group pixels is used as spectral analysis in WiPE algorithm. New algorithm of WiPE is validated in independent dataset and manual checking based on Google map.

Dat Dinh Ngoc, [email protected], Space Technology Institute, Vietnam Academy of Science and Technology, 18 Hoang Quoc Viet, Cau Giay, Hanoi, Vietnam

Duong Dang Nguyen Hien, [email protected], University of Science and Technology of Hanoi, Vietnam Academy of Science and Technology, 18 Hoang Quoc Viet, Cau Giay, Hanoi, Vietnam

Hubert Loisel, [email protected], Laboratoire d’Océanologie et de Géosciences, Université du Littoral-Côte-d’Opale, Université Lille, CNRS, UMR 8187, LOG, 32 avenue Foch, Wimereux, France

Cédric Jamet, [email protected], Laboratoire d’Océanologie et de Géosciences, Université du Littoral-Côte-d’Opale, Université Lille, CNRS, UMR 8187, LOG, 32 avenue Foch, Wimereux, France

Vincent Vantrepotte, [email protected], Laboratoire d’Océanologie et de Géosciences, Université du Littoral-Côte-d’Opale, Université Lille, CNRS, UMR 8187, LOG, 32 avenue Foch, Wimereux, France

Lucile Duforêt-Gaurier, [email protected], Laboratoire d’Océanologie et de Géosciences, Université du Littoral-Côte-d’Opale, Université Lille, CNRS, UMR 8187, LOG, 32 avenue Foch, Wimereux, France

Chung Doan Minh, [email protected], Space Technology Institute, Vietnam Academy of Science and Technology, 18 Hoang Quoc Viet, Cau Giay, Hanoi, Vietnam