In addition to spectral features, texture is an important spatial feature used in identifying objects or regions of interest in an image. Although texture is relatively easy for human observers to recognize and describe in empirical terms, it has been extremely refractory to precise definition and analysis by digital computers. This paper describes a procedure for extracting some easily computable features for the texture of blocks of digital image data and illustrates the applications of combined textural (spatial) and spectral features for identifying the land use categories of blocks of ERTS MSS (Earth Resources Technology Satellite Multi Spectral Scanner) data. The land use classification algorithm based on textural and spectral features was developed and tested using 614 image blocks of 64*64 resolution cells derived from an ERTS image over the Monterrey Bay area of the California coast line. The algorithm was applied on a training set of 314 blocks and tested on a set of 310 blocks. The overall accuracy of the classifier was found to be 83.5% on seven land use categories


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Combined spectral and spatial processing of ERTS imagery data


    Additional title:

    Kombinierte spektrale und raeumliche Verarbeitung von ERTS-Abbildungsdaten


    Contributors:

    Published in:

    Publication date :

    1974


    Size :

    , 20 Quellen



    Type of media :

    Article (Journal)


    Type of material :

    Print


    Language :

    English




    Combined spectral and spatial processing of ERTS imagery data

    Shanmugam, K. / Haralick, R. M. | NTRS | 1973


    Spectral and textural processing of ERTS imagery

    Haralick, R. M. / Bosley, R. J. | NTRS | 1974



    Crop identification using ERTS imagery

    Horton, M. L. / Heilman, J. L. | NTRS | 1973


    Pseudocolor transformation of ERTS imagery

    Lamar, J. V. / Merifield, P. M. | NTRS | 1973