Research On Automatic Extraction Of Coastline From GF1 Remote Sensing Images Based On Object-oriented

The coastal zone is one of the regions with large population aggregation and activity(where is almost 60%people lived).In recent years,human activities along with some natural factors cause a series of ecological environment problems in coastal zones,such as wetland degradation,function decline of ecological service,bio-diversity are destroyed and so on.In this article,feature information and information extraction theory of high resolution remote sensing image information extraction theory and Object oriented classification is firstly introduced,using the types of membership function in the object oriented classification methods,the artificial shorelines and natural shorelines are extracted by combining the features of spectral,shape and context features based on the GF-1 remote sensing image of Bohai Bay,Qinhuangdao,Tangshan,Tianjin and Cangzhou.Aiming at problems those lack of tide correction on extraction of coastline at present,correcting the coastline using model of tide which coastlines are worst effected by tide and get real coastline.Mean while,according to the characteristic information of the high resolution image,the salt and pepper phenomenon can be avoided effectively.Analyzed the accuracy of extracted coastlines using "908" aerial survey coastline and Google Earth Map Calculate the root mean square error and among the results the root mean square error of the sandy shoreline is up to 1.70m The worst quality is muddy coastline and it reached 34.79m the root mean square error of artificial shoreline is generally maintained at about 2m,the breeding shoreline error is slightly larger than 5.17m.Results show the coastlines information extraction method based on object-oriented in this article achieved some good extraction results and provide technical support for scientific protection and rational development of coastal zone.