Aperiodic Digital Halftoning and Its Multitoning Extension

• Main Authors: Jia-Yu Chang, 張嘉裕
• Other Authors: Jing-Ming Guo
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/89123990395977219781

碩士 === 國立臺灣科技大學 === 電機工程系 === 101 === In this thesis, three techniques, including two improved digital halftoning algorithms and one digital multitoning algorithm, are proposed to convert images for output devices. Digital halftoning is a technique for converting continuous-tone images into limited-tone images. The distortion between an original image and the converted image is inevitable, since the number of the corresponding output levels is less than the continuous-tone. Thus, the main object of the halftoning is to reduce the distortion induced from this transformation. Nowadays, the processing performance of the halftoning has become a critical issue as the image resolution is increasing. Another key element of this thesis is to effectively boost the processing efficiency. The first proposed method is the Multiple Look-Up Table (MLUT) which is based on the look-up table halftoning, and is an effective technique for yielding satfactory image quality. According to the experiment results, the dot distribution generated by the proposed method can approximate to the well-known direct binary search halftoning which can achieve the best image quality so far, and thus it can be a very competitive candidate in coping printing industry. The second proposed method is the Content-Dependent Dot Diffusion (CDDD) which is based on Dot Diffusion (DD). The proposed CDDD can provide better image quality and a near aperiodic characteristic simultaneously comparing to the former parallel methods. The multitoning is a technique which extends the halftoning by adopting more than two quantification levels for reducing the distortion between an original image and the converted image. Yet, the banding effect disturbs the visual perception, and thus degrades the image quality. To improve the image quality by removing the banding effect, the third method termed Tone-Replacement Error Diffusion multitoning (M-TRED), is proposed. According to the experimental results, the proposed method can provide excellent tone-similarity and dot-distribution simultaneously comparing to the former banding-free methods in the literature.