Optimized Inverse Halftoning Technique Based on Bilateral Filter

Abstract

Halftoning and inverse-halftoning processes are tools that have been widely used in the last few years, which are used in scientific research for the development of visual cryptography, steganography, reconstruction of tampered images, authentication systems, among others. In this paper, a technique of inverse-halftoning, which generates grayscale images from halftone images, is proposed. This method obtains a high quality in the generated grayscale images. Our approach first uses a low-complexity low-pass filter to obtain an approximation to the desired grayscale image, and then a bilateral filter is applied in order to eliminate regions with noise without affecting the edges of the image. To evaluate the quality of the obtained images, the objective criterion of peak signal-to-noise ratio (PSNR) is used. The experimental results demonstrate the effectiveness of our method in comparisons with other schemes reported in the literature, additionally, the average processing time for the proposed system maintains a good performance against the reported schemes.