Research On Sea Ice Retrieval Algorithm Based On Satellite Remote Sensing Data

As an important maritime hub,the Bohai Bay provides great convenience for shipping,and at the same time suffers from sea ice disasters of different severity every winter,and the socio-economic and development of the region has been greatly affected.At present,the research on the inversion of the Bohai Sea ice information based on the Fengyun Meteorological Satellite FY-4A independently developed by my country is still insufficient.Therefore,this paper uses FY-4A data to study the Bohai Sea ice inversion,realizes the Bohai Sea ice identification,and provides data support and algorithm support for ocean climate forecasting related departments.The main research work of this paper is as follows:(1)The Bohai Sea ice inversion algorithm based on threshold and FY-4A data.Explains the basic principles of remote sensing monitoring and sea ice inversion.After data preprocessing,the data is first processed for land removal and cloud detection.The Bohai Sea ice is inverted using the multi-channel fixed threshold method,and then combined with multichannel thresholds.The method and adaptive threshold algorithm realize the inversion of Bohai Sea ice under clear sky conditions.Experiments show that under clear sky conditions,FY-4A data has certain usability for Bohai Sea ice inversion.(2)Integrated inversion of Bohai Sea ice based on random forest.The classification algorithm based on random forest is studied,combined with the sea ice inversion method,the random forest classification algorithm is introduced into the sea ice inversion,which can achieve a certain effect of classification inversion for the sea ice under the cloud cover.Then the seed filling algorithm is introduced to correct the results of sea ice inversion and realize hole filling.Experiments show that under non-clear sky conditions,the results of the Bohai Sea ice integrated inversion based on random forests have been improved.The proposed algorithm can effectively identify the Bohai Sea ice and improve the accuracy of sea ice inversion.The accuracy and stability of the performance laid the foundation.The research results in this paper also lay the foundation for the commercialization of satellite remote sensing by my country’s FY-4A,and the generated results can ensure the real-time monitoring and prediction of sea ice in the Bohai Sea.(3)Implementation of a prototype system of Bohai Sea ice monitoring software based on FY-4A data.After researching the sea ice recognition algorithm,this paper carried out the design and development of the Bohai Sea ice inversion prototype system based on FY-4A data and random forest,and realized the commercialization of sea ice monitoring.