In this study it is aimed to form vectorized drawing files of digital images which are converted from satellite and aerial photos, according to their radiometric properties.When digital images acquired from scanner they are kept in raster form. In raster form the image is expressed by matrices which contain pixel values. When saving an image like this in computer environment, it takes too much disk space and also it is difficult to edit obtained image. Therefore, it is necessary to convert these images to vector form. Therefore they can take much less space and editing become easy. In literature, there are so many works in this area, but none of them gives perfect results. In this study, the works in this topic examined and the two of them are implemented by MATLAB software codes. One of the implemented algorithms is more useful for engineering drawings and keeps structural form of the drawing and uses line tracking. The other method, keeps the structural form, moreover, it keeps the line width and named sparse pixel. In this study a different algorithm developed based this method which keeps line width. Unlike the method taken as base the result vectors are represented as spline curves.Key words : Vectorization, Raster to Vector Conversion, Image Processing, Feature, extraction |