A novel algorithm for inverse halftoning using LUT approach and pattern labeling

Abstract

Due to binary nature of halftone images it is often difficult to process them without causing keen degradation compared to grayscale images. Therefore, it gives the purpose to develop many algorithms to reconstruct grayscale images from their corresponding halftones which is called inverse halftoning. A fast and computation-free technique to accomplish inverse halftoning is the Look-Up Table (LUT) method that leads to obtain images with admissible quality. In this paper a novel technique is introduced in an attempt to improve the LUT approach proposed by Murat's and Vaidaynatan by using pattern labeling. By means of simulation, we show the efficiency of this technique. Also, by calculating average PSNR, we compare its competency with some other known techniques. Furthermore, we discuss about the complexity of implementing this technique in this paper.