Component Development About Vectorization Of Raster

The paper aims to solve the problem,How to change the Raster images to Vector? This is a part of Map Pattern Recognition."Raster to Vector" recognizes line in the rastor map to a chain of vertex coordinates,creates a line of vector type finally.Map Pattern Recognition is a subject of synthetic technology,involves Computer science, Digital image processing,Computer vision,Pattern recognization Artificial intelligence and so on.It recognizes and extracts useful information from map,utilized those subjects 'technologies referring.Geological exploration of the crisis mines accumulates much information,range from prospecting to supplemental work after exploration,including geological,geophysical, geochemical,remote sensing,and deep prospecting data.Besides,this information is added continuously through work in replacement resources exploration of crisis mines.So,various maps became the expressions of products inevitably,image vector to geological compilation map and other objects is a useful and convenient method for information acquisition.The paper referes to Spatial Data Model,that consists of rastor and vector type.In rastor type,it underlines to some forms of image(BMP,PCX,GIF,JPEG,JPEG2000,PNG,TIFF, GEOTIFF).In vector type,it emphasizes on ESRI ShapeFile.The next,it specifys uniform color space.In this space,the value of every color range accords with the value of true color we know.The color difference formula consists of three formula,CIE 1976 Lab color difference,CMC(l:c) color difference,CIE DE2000 color difference.Image threshold alms to get valuable information in map,so we can carry on succedent process,and extract important information.The thinning arithmetic is a part of Mathematical Morphology,using Morphology process image.Image vectorization gets vector coordinate chain from thinned line.Vectorization arithmetic is Douglas Peucker.After finishing vector,we maybe carry on map projection,because input map is a projection,output map is another projection.Using the method of Affine Transformation,Similarity Transformation,or Projective Transformation, we can transform from digitized map coordinate to true coordinate(e.g.UTM or Gauss).In conclusion,using process of picking up thic line,thining,vectorization,sparsity,can recognize a near color,less than some width value,connected line.