Two researchers from Microsoft and the Hebrew University of Jerusalem, just presented a method of image processing announced as revolutionary. This is a method to recalculate the pixels of an image type 8-bit to get out an HD version automatically by the machine completely redesigned.
Specifically, it involves transforming a raster image to give it a render HD via an algorithm that will calculate the rendering of the new image from the pixels of the original low resolution. The algorithm, created by Johannes Kopf and Dani Lischinski, identifies the definition of the image, calculates the color of each pixel and recreates the shadows, contours and other lines making up the drawing.
The few examples presented elsewhere speak for themselves. Note however that the algorithm works only with a very specific type of images for the moment and is unable, for example, to give better definition to a photo.
No comments:
Post a Comment
Note: only a member of this blog may post a comment.