Abstract Detection of different regions like impervious surfaces, vegetation and water from a multispectral satellite image is a complex task. This paper introduces a novel idea for impervious surface detection from multispectral images using SURF descriptors. To determine the efficiency of the proposed system, a comparative evaluation is done with other two techniques, namely histogram based and spectral-value-based technique. The result shows that the proposed system outperforms the other two techniques in detecting impervious surfaces like buildings and vehicles with an accuracy of 80.48%. The histogram-based technique and spectral-value-based clustering obtained an accuracy of 61.89% and 68.29% respectively. However, in classifying vegetation the other two techniques outperforms SURF descriptors. The histogram based technique gives an accuracy of 86.46% and an accuracy of 94.35% is obtained by using the spectral-value-based clustering. Whereas SURF based technique gives only an accuracy of 50.71%.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Impervious Surface Detection from Multispectral Images Using Surf

    Paulose, A. / Sreeraj, M. / Harikrishnan, V. | British Library Conference Proceedings | 2014


    Air-impervious liner contour roll and air-impervious liner

    ZHANG MIN / YANG DAWEI / ZHANG YUNFEI et al. | European Patent Office | 2020

    Free access

    Vehicle speed estimation using extracted SURF features from stereo images

    El Bouziady, Abderrahim / Thami, Rachid Oulad Haj / Ghogho, Mounir et al. | IEEE | 2018


    Cloud Detection for MERIS Multispectral Images

    Gomez-Chova, L. / Camps-Valls, G. / Amoros, J. et al. | British Library Conference Proceedings | 2005


    Edge detection in multispectral remote sensing images

    Sirin, T. / Saglam, M.I. / Erer, I. et al. | IEEE | 2005