Inverse halftoning using Projection Onto Convex Sets (POCS) by interval consistency check method for error diffused images
Abstract
Halftoning breaks the continuous tone image into a pattern of black dots and makes an illusion of a continuous image to the observer. Halftoned image is a binary version of the continuous toned image. Halftoning techniques have been widely accepted for the printing of newspapers, magazines as well as fax machines and printers. Inverse halftoning is the reconstruction of continuous toned image from its halftoned version. The need for inverse halftoning arises from the need to manipulate the images, rotation or zooming etc In this thesis we deal with inverse halftoning of images halftoned by error diffusion algorithms . Analysing the spectrum the low frequencies of the halftoned and original image are similar, but the halftoned image contains quantization noise in the higher frequencies. So the first step in our inverse halftoning method is the removal of high frequency quantization noise by Gaussian lowpass filtering. The second stage of our inverse halftoning algorithm is the Projection Onto Convex Sets (POCS) algorithm which alternatively projects the image onto the two convex sets, repeats and converges in the intersection of the two convex sets. (Abstract shortened by UMI.)
Subject Area
Electrical engineering
Recommended Citation
Iyer, Krishnan, "Inverse halftoning using Projection Onto Convex Sets (POCS) by interval consistency check method for error diffused images" (2005). ETD Collection for University of Texas, El Paso. AAI1427716.
https://scholarworks.utep.edu/dissertations/AAI1427716