Orientation field guided line abstraction for 3D printing

This paper focuses on printing a 2D color image by using a desktop fused deposition modeling (FDM) 3D printer. A fully automatic framework is presented to convert a 2D image into a non-photorealistic line drawing which is suitable for 3D printing. Firstly, an image is partitioned into a moderate number of regions, and contours of these regions are extracted to deliver the high level abstraction of the image. The contour lines are further refined by taking into consideration the restrictions of 3D printing. Next, an orientation field based on the contours and feature lines at a finer level is computed to guide the placement of streamlines. The distances between streamlines are carefully controlled such that the density respects the pixel intensity values. Finally, the resulting streamlines are converted into printing paths and printed by using filaments with specified colors. Experimental results show the feasibility and efficacy of our method on portraying a given image by using a few of 3D printable non-intersecting lines while preserving features and tone variation in the image.