Vectorization Design For Topographic Map And The Application In Three-dimensional Reconstruction

Image vectorization technique is a kind of important image processing technology. It is also a multi-scale technology which involves graphics and some knowledge in other subjects. Especially in recent years, with the rapid development in computer CAD drawing and geographic information system (GIS) related field, the technology of image vectorization is the key step. The research of image vectorization has important value and commercial value.Due to the different image characteristics, even there has a variety of general image vectorization method; no one can meet all the need of images vectorization with different characteristics. Therefore this paper only specific to a kind of linear image and a kind of map which colors can be clustered. This paper studies such images'vectorization method.For the topographic map which mainly has the linear elements, this paper put forward an image enhancement method which mainly using mathematical morphology. And in the light of the image's line is easy to break and adhere, this paper revise these faults. Aiming at the characteristics of linear image, this paper put forward a vectorization optimization algorithm method based on thinning, using genetic algorithm to find the key points, and using straight and arc as line fitting.For the topographic map which colors can be clustered, compared with other methods, this paper put forward the use of FCM color clustering algorithm to color clustering in image regions. The experimental results show that this method can keep the topological relationships between regions in the original image well, and received good effect. This paper describes regional topology relation region used the list and then gets the contour skeleton image. Finally this paper used the forgoing method to vector the image.Aiming at the characteristics of vector data, this paper used SVG formats for vector standard image data storage, using this kind of SVG formats has the advantages of image vector changes and interact with GIS software.Using the generated SVG vector contour map, and combining with the improved interpolation algorithms, we can get the approximate elevation of grid points. Then we can use the GIS tools for 3D reconstruction with these topographic maps. Through the experiments, the improved interpolation algorithm has high accuracy, and reconstruction reaches the expected effect.Overall, using the kind of method of image vectorization put forward in this paper, the characteristics of digital image in these images can be held very well. And this method provides the convenience for scanning images'digitization.