METHOD FOR DECREASE OF IMPACT OF NOISY HYPER SPECTRAL CHARACTERISTICS OF REFLECTION 
OF POLLUTED SEA WATER 

Eminov Ramiz Akhmed ogly, candidate of technical sciences, associate professor, ub-department of transportation and storage of oil and gas, Azerbaijan State University of Oil and Industry (20 Azadlig avenue, Baku, Azerbaijan), Eminov.ramiz@mail.ru
Ismailov Magerram Magomed ogly, postgraduate student, Azerbaijan State University of Oil and Industry (20 Azadlig avenue, Baku, Azerbaijan), ismailovmm@mail.ru
Asadov Hikmat Hamid oglu, doctor of technical sciences, professor, head of department of research institute of aerospace informatics, National Aerospace Agency (1 S.S. Akhundov street, Baku, Azerbaijan), asadzade@rambler.ru
Ismailov Kamal Kheyraddin oglu, doctor of technical sciences, associate professor, National Academy of Aviation (25-th km, Bina, Baku, Azerbaijan), kamal.ismaylov@mail.ru

551.463

10.21685/2307-5538-2018-1-8

Background. The high livel of pollution of sea waters in zones of sea oil production necessitates the carrying out of continual ecological control of sea waters condition. The control of sea waters condition is carried out by onboard spectroradiometric instruments installed on the board of carriers and for analysis of derived data the methods of derivative analysis is broadly used. In this method the feature taken into account is that the first derivative stresses out not only the part of noises, but also the specific spectral features of studied object. 
Materials and methods. The method making it possible to carry out the preprocessing of hyper spectral signal using the first derivative in condition ühere the noisy components of the initial measuring signal is prevalent has been suggested.
Results. The pecularity of suggested method is that it make possible to define the function of representation of first derivative on the basis of computed initial spectral characteristics which is optimum on criterion achieving the extremum by conditional information content of hyperspectral data. Conclusions. The suggested method could be used in the technics of remote sensing of sea water condition to decrease the effect of noisy component of hyper spectral data obtained using the hyper spectral instruments.

hyper spectrometer, information, derivative characteristic, noisy signal, remote sensing