Method of sparse front for vectorization of lined images
AbstractA new method of vectorization of lned images is proposed. It is based on the algorithm of sparsely-pixel tracking straight and curved lines on the bitmap. The result of this algorithm is the set of trajectories of lines in the form of sequences of points. The novelty of the method is to use weights while calculating the points of the trajectories that would reduce the dependence of results of vectorization from noise contours lines on the bitmap. Also an efficient algorithm of counteraction to re-tracing the line of the present method is proposed. At the second, the final stage of vectorization obtained trajectories are transformed into a set of vector primitives such as lines and arcs, the combination of which approximates straight and curved lines and forms a vector image. The algorithm has a high performance and can operate without settings. Comparative research of the performance of the algorithm and the quality of the results of its work is conducted.
Janssen R.D.T., Vossepoel A.M. Adaptative vectorization of line drawing images // Computer vision and image understanding. — 1997. — 65. — P. 38–56.
Wenyin L., Dori D. From Raster to Vectors: Extracting Visual Information from Line Drawings // Pattern Analysis and Applications. — 1999. — 2, N 1. — P.10–21.
Pomekhoustoychivyy volnovoy algoritm vektorizatsii lineynykh rastrovykh ob”yektov. — http://stanislavmoskalenko.narod.ru/articles/article_mashin3.htm.
Dori D., Liu W. Sparse Pixel Vectorization: An Algorithm and Its Performance Evaluation // IEEE Pattern Analysis and Machine Intelligence. — 1999. — 21, N 3. — P. 202–215.
Song J., Su F., Chen J., Tai C., Cai S. Line Net Global Vectorization: an Algorithm and Its Performance Evaluation // Computer Vision and Pattern Recognition 2000. — 2000. — P. 1383–1388.
OpenCV. — http://opencv.org.
OpenCV Hough Line Transform. — http://docs.opencv.org/doc/tutorials/imgproc/imgtrans/hough_lines/hough_lines.html.
Stephens R.S. A Probabilistic Approach to the Hough Transform // Image and Vision Computing. — 1991. — 9, N 1. — P. 66–71.
Bradley D., Roth G. Adaptive Thresholding Using the Integral Image // ACM Journal of Graphics Tools. — 2007. — 12, N 2. — P. 13–21.