Sovremennye problemy distantsionnogo zondirovaniya Zemli iz kosmosa, 2012, Vol. 9, No. 5, pp. 301-311
A lay-out of the apparatus and programmatic complex of hyperspectral airborne remote sensing data processing
T.V. Kondranin
1, V.V. Kozoderov
2, E.V. Dmitriev
3, V.D. Egorov
3, V.V. Borzyak
21 Moscow Institute for Physics and Technology (State University)
2 M.V. Lomonosov Moscow State University
3 Institute of Numerical Mathematics of Russian Academy of Sciences
Computational procedures of natural and anthropogenic objects recognition with supervising using test samples give a basis of the created lay-out of the apparatus and programmatic complex. Characteristic features of one of the most perspective approaches are outlined to solve the recognition problem for such objects while applying hyperspectral data with the utilization of statistical techniques of discriminant analysis that ensures the minimal mean square error of classification (the Bayesian approach). The usage of a step up method that enables to conduct the selection of the most informative features with the optimization of the applied remote sensing channels is realized as a result of the proposed approach.
Keywords: hyperspectral airborne remote sensing, data processing, apparatus and programmatic complex, recognition of natural and anthropogenic objects
Full textReferences:
- Kondranin T.V., Kozoderov V.V., Kazantsev O.Yu., Bobylev V.I., Dmitriev E.V., Egorov V.D., Ka¬mentsev V.P., Borzyak V.V., Sovremennye problemy distantsionnogo zondirovaniya Zemli iz kosmosa, 2011, Issue 8, Vol. 1, pp. 90-98.
- Kozoderov V.V., Dmitriev E.V., Egorov V.D., Borzyak V.V., Sovremennye problemy distantsionnogo zondirovaniya Zemli iz kosmosa, 2011, Issue 8, Vol. 3, pp. 55-63.
- Kozoderov V.V., Dmitriev E.V., Vestnik Moskovskogo gosudarstvennogo universiteta lesa, 2012, No.1.