Johannes Kopf (Microsoft Research) and Dani Lischinski (The Hebrew University) have developed a new way to make vector art from pixelated art.  Pixel art was has been used since the first video games were made. It worked great when the screens being used had very low resolution.  Pixel art has the downside that when you try to enlarge the image it becomes blurry.  The nearest neighbor interpolation algorith allows  you to smooth out the blurriness but it still has it’s limits.  Now Vector graphics, on the other hand, uses lines, curves and points that are scaled to each other.  This allows you to enlarge or shrink a picture without any loss of resolution.

We describe a novel algorithm for extracting a resolution-independent vector representation from pixel art images, which enables magnifying the results by an arbitrary amount without image degradation. Our algorithm resolves pixel-scale features in the input and converts them into regions with smoothly varying shading that are crisply separated by piecewise-smooth contour curves. In the original image, pixels are represented on a square pixel lattice, where diagonal neighbors are only connected through a single point. This causes thin features to become visually disconnected under magnification by conventional means, and it causes connectedness and separation of diagonal neighbors to be ambiguous. The key to our algorithm is in resolving these ambiguities. This enables us to reshape the pixel cells so that neighboring pixels belonging to the same feature are connected through edges, thereby preserving the feature connectivity under magnification. We reduce pixel aliasing artifacts and improve smoothness by fitting spline curves to contours in the image and optimizing their control points.

Original Super Mario World Dolphin (SNES)
Super Mario World Dolphin after Vector Transformation

Quite an amazing difference. While the paper doesn’t go into specific applications for this, I can definitely see the potential for updating the graphics on older system for modern televisions.  Check out this link for the full paper and watch the video below for a side by side comparison of the algorithm.

[flashvideo file= width=545 height=307 /]

Share this post: