Research On Detection And Identification Technology Of Oil Spill At Sea Surface By Electromagnetic Remote Sensing

About 71% of the earth’s surface is covered by blue ocean.With the development of the global economy,people are paying more and more attention to the value of the ocean.While developing and utilizing the ocean,the Marine ecological environment is being continuously damaged,such as frequent oil spill accidents.It is of great scientific significance to carry out high-precision monitoring to protect the Marine environment after the sudden oil spill accident.With the continuous development of global satellite remote sensing technology,electromagnetic remote sensing detection and identification technology has become an important means to monitor oil film on the sea surface.In this paper,the Marine oil spill accident is taken as the research object to study the physical mechanism,identification technology and three kinds of optical monitoring classification models of Marine oil film by SAR(synthetic aperture radar)remote sensing monitoring.The main research results are as follows:(1)physical mechanism of SAR remote sensing monitoring of oil spill on the sea surface.Based on the electromagnetic scattering model,the scattering of radar electromagnetic wave signal is described,and the concept of equivalent relative dielectric constant of oil film covering sea surface is proposed.Based on the equivalent relative permittivity model of oil film,the mapping function between the dielectric polarization ratio of thin film and the equivalent relative permittivity is numerically studied,and the polarization ratio model can be used to distinguish oil film from seawater.(2)study on SAR remote sensing monitoring and recognition algorithm for oil spill on the sea surface.Based on the study of the physical mechanism of surface oil spill SAR monitoring,a method of surface oil spill identification using SAR polarization ratio and texture characteristics was proposed for oil film identification experiments,and the results of VV single-polarization image and polarization ratio image identification were further analyzed.The experimental results show that: based on the polarization ratio image texture feature method to identify the oil spill information can well realize the monitoring of the oil spill area,reduce the misjudgment caused by the similar oil spill phenomenon in the monitoring process,reduce the false alarm of oil spill identification by 17.96%,and achieve the overall oil spill identification accuracy of 96.83%.(3)an automatic monitoring and treatment system for Marine oil spill has been developed.Based on the proposed oil spill identification method,a python language microwave remote sensing monitoring system for Marine oil spill was developed.The system integrates the whole set of sea surface oil spill business detection process,including radarsat-2’s XML file reading,data preprocessing,texture feature analysis and oil oil extraction functions,making the whole monitoring business process more efficient,simple and efficient.(4)three optical surveillance classification methods(parallel pipeline method,maximum likelihood method and support vector machine method)for oil film recognition were evaluated.Based on the overall accuracy and Kappa coefficient,the ability of different classification methods to identify oil film was analyzed,and the applicable scope of each classification method for the classification of TM oil spill images in this paper was summarized.The experimental results show that SVM has the highest accuracy compared with the other two classification methods.