Line Drawing For 3D Printing

In the research field of computer graphics and vision,non-photorealistic rending(NPR)has become a hot topic in recent years,which requires using the computer to generate graphics with hand-painted style.Stippling and line drawing as one of NPR performance,they aim to use simple geometric elements such as points and lines to capture the characteristics of image,for example,they use lines to portray contour feature,or use the density of points or lines to reflect the tone and texture etc.Therefore designing a simple,natural and efficient stippling and line drawing have been a research focus in computer graphics for a long time.Aim at 3D printing application,this paper puts forward a line drawing method of keeping image features.The method uses the watershed algorithm for image segmentation in the first place,then regards the segmentation boundary as the feature line and extract it.Then we use the Lloyd iterations to relax the points,during the process we remain the points on the feature line,and the interior points move to the center of voronoi cells.After the completion of the iterations,we resample the feature line and getting the stippling of keeping image feature.Based on the result of stippling,the problem can be converted to TSP problem,and we use LKH algorithm to solve in block efficiently.Comparing with the previous methods,our method designs for 3D printing.With the promotion of 3D printing,people can make use of 3D printers to printing items personally that oneself like.As everyone knows,3D printer is widely applied to print a 3D model,but applying 3D printer to print image rarely involved.And the general line drawing methods are not suitable for 3D printing,this is because they do not consider the printer requirements,their lines may be intersecting,too dense,or the line segments may be too many,these will lead to wiredrawing,nozzle extrusion etc.,and affect the appearance of printing result.In the end,this paper shows the comparison ofour method with other popular line drawing techniques,including edge detection,TSP art,vector field visualization and weighted line drawing.From the perspective of line drawing and 3D printing result,our method outperforms others in that it is not only capable of capturing image features,but also capable of depicting them with smooth and coherent lines and suitable for 3D printing.