Road extraction from very high-resolution (VHR) true color (RGB) images is a popular task in the area of image processing and computer vision. However, there are a number of obstacles that make this task difficult, such as road coverage by tree canopies, shadows, and similar surface objects. To fine-tune the task of road feature extraction, we have proposed a hybrid approach involving various popular image processing functions such as thresholding, fusion, and morphology. Moreover, road extraction from satellite-based VHR-RGB images is quite challenging as compared to doing the same task using multispectral or hyperspectral images. After road extraction, we have also analyzed the efficacy of our proposed approach using several machine learning algorithms. Ultimately, a higher performance with 91.52% accuracy was achieved for the task using the proposed approach.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Chromatic Logical Fusion for Road Feature Extraction in VHR-RGB Satellite Images


    Beteiligte:


    Erscheinungsdatum :

    23.02.2024


    Format / Umfang :

    1673701 byte




    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Dehazing and Road Feature Extraction from Satellite Images

    Gopan, Archa / Muhammed, Abid Hussain | IEEE | 2019


    Automatic Road Extraction using Scale Invariant Feature Transform-based Random Forest in Satellite Images

    Rajesh, N / Alzubaidi, Laith H. / Shivaprasad Yadav, S. G. et al. | IEEE | 2024