Raster Image Boundary Vectorization Program

Vector graphics have a number of advantages over raster. The process of converting from a raster representation to a vector is called image tracing or vectorization. To achieve the best result, the approach to vectorization of the image should be chosen based on the type of image. This paper compares some vectorization methods as well as the general concept of tracing images. The vectorization method based on human perception and its shortcomings are described in detail in the article. The method aims to vectorize images with distinctly colored closed areas. The developed program is a web application for the automatic vectorization of images involving machine learning. Applied algorithm vectorization is an improved method based on human perception. The paper proposes a method of segmentation by color. This development allows vectorization with saving the result on the server. The implementation provides the ability to configure the process of tracing a bitmap, removing some steps from it and changing the threshold values. This paper contains 27 pages, 3 chapters, 16 illustrations, 11 bibliography items, 8 appendices. Keywords: vector graphics, image vectorization, raster graphics, human perception