Study On Dynamic Monitoring Of Coastline Based On Satellite Remote Sensing

Coastline survey based on water information extraction of remote sensing images is one of the main methods of coastline acquisition.Although the traditional water remote sensing information extraction models have been widely used and continuously improved,there are still some problems,such as misjudging buildings and shadows as water bodies,and misjudging water bodies with high suspended sediment content as non-water bodies.This leads to poor integrity and low accuracy of the extraction results,and it is difficult to meet the needs of high-precision extraction of remote sensing information in a wide range of water bodies.To solve this problem,this paper proposes a method of coastline remote sensing information extraction based on tasselled cap transformation.Firstly,compressing the multispectral image into a new cap transform space by tasselledcap transform.Based on this characteristic space,the difference between the wetness component and the greenness component of oceanic water,coastal water,lake and other surface features is clarified,and the water extraction model was established.Secondly,on the basis of water body information extraction,space vectorization is carried out to obtain coastline information.And the accuracy of the model is evaluated from both qualitative and quantitative aspects.Finally,Zhoushan City and its adjacent areas are selected as the study area to carry out the analysis of the temporal and spatial evolution of the coastline.The experimental results show that this method can effectively eliminate the influence of buildings,shadows and suspended sediment on the extraction results.It has higher Kappa coefficient(0.91),overall accuracy(97.02%)and user accuracy(90.81%)than other mainstream water extraction methods.From 2000 to 2018,the coastline of Zhoushan City changed significantly.The total length of the coastline increased by about205.30 km,the average growth rate was 1.59%,and the land area increased by about124.66 km~2,the average growth rate was 1.57%.From 2000 to 2009,the growth trend of coastline length was lower than average,but it changed greatly from 2009 to 2012,The length of coastline increased to 138.24 km,2.16%higher than that of 2009,and the land area increased to 51.8 km~2,3.97%higher than that of 2009.The growth rate of coastline length declined between 2012 and 2015,but was significantly higher than that before 2009.From 2015 to 2018,although the growth rate of shoreline became lower,the land area increased significantly.The main growth points are Cengang,Donggang,Baiquan and Beichan of Zhoushan Island,Changzhi Island,Zhujiajian and a series of surrounding larger islands,as well as Shengsi,Daishan and other areas.In this study,a new coastline remote sensing information extraction model is established based on the tasselled cap transformation,which makes up for the shortcomings of traditional extraction methods,improves the accuracy of automatic extraction model,and provides technical support for coastline management and government decision-making.