一种基于LUT的图像逆半调改进算法

An Improved Image Inverse Halftoning Algorithm Based on LUT

为提高基于LUT(查找表)的图像逆半调传统方法在空值估计值与逆半调映射值的准确性、提高逆半调图像的质量与感观效果,结合已有的数字图像半调与逆半调技术,提出了一种多域值分块与动态估值优化算法。该算法采用二分迭代多域值分块法与8邻域均值动态递归估值法,通过模板选择、初始化LUT表、LUT逆半调和完整LUT表对传统的LUT算法进行改进与优化。实验表明利用该优化算法进行图像逆半调,能够得到比传统LUT逆半调较好的视觉效果,并且具有较快的运行速度。

In order to improve the accuracy of the null value estimation and inverse halftoning mapping value of traditional inverse half- toning methods based on LUT, enhance the inverse halftoning image quality and sensory effects, combined with a existing digital image halftoning and inverse halftoning technology, propose a multi-domain value of the sub-block optimization algorithm with dynamic evaluation. The algorithm uses a dichotomy iterative multi-domain value block method with 8 neighborhood mean dynamic re, cursive method of evaluation, through template selection,initializing the LUT table, LUT inverse halftoning and complete LUT table to improve and optimize traditional LUT algorithm. The experiments show that the optimization algorithm for inverse halftoning is able to get a better visual effect than traditional LUT inverse halftoning and faster running speed.