형태 정보를 고려한 Halftoning과 Multitiong의 화질 개선에 대한 연구

Abstract

In this thesis, we propose a new halftoning scheme that preserves the structure and tone similarities while maintaining the simplicity of Floyd-Steinberg error diffusion. Our algorithm is based on Floyd-Steinberg error diffusion algorithm, but it is modified in the part of the threshold modulation to improve over-blurring of Floyd-Steinberg error diffusion algorithm. Since our algorithm changes the quantizer thresholds according to the structure information and random noise, it preserves the details in the textured region. Simultaneously, it also reduces visual artifacts of the original error diffusion in the uniform region. Unlike existing error diffusion algorithms, the proposed algorithm preserves fine texture very well even in low contrast images. Our algorithm has been tested on various types of images including low contrast images and shows visually pleasing results both in the textured region, and in the uniform region.Plasma Display Panels (PDPs) suffer from motion artifacts such as motion picture distortion (MPD). To reduce MPD, PDPs use the limited gray levels instead of the whole 256 gray levels. Therefore we need multitoning to represent the absent gray levels among the whole gray levels. We extend our algorithm for multitoning in PDP.