Research Of Salinity Information Extraction Based On Ooc

Soil salinization not only causes the destruction of resources and immense decline of agricultural productivity, but also threatening the biosphere and the constitution of ecological environment. Soil salination often occurs in these area,where the climate is drought,the soil's evaporation is very intensity,and the water table is high and contain high dissolubility salt.It was occurred by certain climate, terrain, and hydrogeology, and they together influenced the water and salt movement. At the present time, soil salinization and secondary salinization have been an basilic issue of entironment ,that influence the production of industrial and agricultural seriously. Then comprehending these factors by the way of general-purpose,has a great sense of making reasonable measures to develop the governance.The t raditional classification and ext raction of the information about remote sensing mainlymake use of the method combining science statistic with man-made interpretation , which is low inprecision and inefficient . Meanwhile , the method leaning on the people who participate in theinterpretation and analysis doesn't have repetition to some extent .the classification technology based on object is different f rom thepure spect ral information classification , the image object also includes many other features used toclassify information such as shape , vein and interrelation and so on. A comparison of the classificationresult s of eCognition and the t raditional classification method indicates that the feature ext ractionoperator of eCognition is more adapt for auto-identify classification of high-resolution remote sensingwhose information is abundant in st ructure and geomet ry than the t raditional classification method.(1)This article used the ALOS image years of 2007 , geometric correction and terra feflaction.obtained the vegetation index and soil salinity index,must be use the terra feflaction,so make use of head fileALOS imag and formula receive the surface reflective image .(2)Use of command Featrue Extraction of ENVI ZOOM , Featrue Extraction for vegetation indx and soil index ,then save the vector image , by then prossed density analysis,geositatis analysis in ArcGIS.result of density analysis image of vegetation index dark color and high density erea is distribution of vegetation comparativly much.and iamge of soil index this playce erea of lightnes salinity destiny lower.mesne of siol index imege lihgt color ,upper density erea is the distribution a fat lot,moreover this playsis is high salinity erea.geositatis analysis result apprent the vegetation index of this erea is not the corridation distrbution ,salinity index is belong tu the corridation distribution.these results for the latter of the clasfication and effective classification affored for the thereunder and reference.(3) image segmentation has been processed on the remote sensing image using the method of multi-scale segmentation method,try the different segment parameter 10,30,50,80.scale parameter exceed biness,some detail information is lost,and directness touch the result of classification.trial and error indicate result scal paramater 30be propitious to extrection of salinization information.(4)First for processed classification the maximum likelihood ,classification mattrix for the86.68%.choose scal parameter for 30 then processed the classification.,classification matrix is92.26. The results showed that: the use of object-oriented method on the classification of ALOS image, not only reduce the"salt and pepper phenomenon"effectively, but also it has the higher classification accuracy than traditional classification mothods, and will provide a broader prospect to the automatic extraction of saline soil information .