Research On Spatial Distribution And Extraction Of Color Steel Sheds Based On High-resolution Remote Sensing Imagery

Color steel sheds have become one of the signs of rapid urban construction because they are constituted by special material.Now they are widely distributed in industrial parks,new technology development zones,urban villages,joint of urban and rural areas,urban edge and other urban plots.Although color steel sheds promote urban development,they bring harm to city.For example,there are more fire disasters,more illegal buildings,more environmental pollutions in these areas where a large number of temporary color steel sheds gather.These areas are vulnerable and need to be monitored.High-resolution remote sensing imagery can effectively detect color steel sheds on the ground.But according to the study of literatures,there is no report about color steel sheds extraction from remote sensing imagery.Domestic high-resolution satellite imagery have the advantages of precision,time and cost,so this paper used them as data source.Firstly,color steel sheds extraction from GF-1 was studied;secondly,spatial distribution of color steel sheds was analyzed;finally,the relation between spatial distribution of them and urban spatial structure was discussed.The main research results are as follows :(1)An extraction model of color steel sheds based on object-oriented was proposed.Because of the disparity of size of color steel sheds,so the better segmentation scales consisted of a smaller initial segmentation scale and a larger block merging scale.Then the optimal segmentation scale can be selected from the better segmentation scales.At optimal segmentation scale,both the largest color steel shed and the smallest color steel shed in the imagery should be extracted.In fact,the segmentation scale,at which the ratio of the maximum area and the minimum area of color steel sheds extracted properly is the largest,is the optimal segmentation scale.Target feature extraction is an important step after imagery segmentation.This paper excavated characteristics of color,spectrum and space of color steel sheds,and constructed decision tree model based on knowledge rule of color steel shed s.The experimental results show that this method can effectively extract color steel shed s from imagery.(2)The spatial distribution of color steel sheds was studied.Color steel sheds in Anning district were selected as research object.Average nearest neighbor index shows that small and large color steel sheds are all aggregated,the aggregation of small color steel sheds is particularly evident;Standard deviational ellipse is used to analyze distribution direction of color steel sheds,the results show that their distribution direction is southeast-northwest axist,and large color steel sheds’ diffusion trend is more significant;Kernel density analysis shows that small color steel sheds are concentrated in urban sub-center,while large color steel sheds are concentrated in urban sub-fringe,small color steel sheds and large color steel sheds have a tendency to be distributed along both sides of road in urban fringe.(3)The relation between spatial distribution of color steel sheds and urban spatial structure was studied.Correlation analysis indicates that there is a moderately positive correlation between total area of small color steel sheds and agricultural population,highly positive correlation between total area of large color steel sheds and the number of enterprises above designated size,moderately positive correlation between total area of large color steel sheds and output of enterprises above designated size;Semivariogram was used to analyze spatial autocorrelation of color steel sheds,and results reflect that city has been expanding gradually along the long axis;Boundary of color steel sheds based on voronoi diagram was studied,results show that boundary of color steel sheds is consistent with the boundary of urban built-up area.