基于成像模型提高着色真实性

Improving realism of colorization based on imaging model

针对现有的着色算法在图像的高光或暗影区难以获得满意效果的问题,通过分析自然彩色图像的成像原理,提出了一种后处理方法来改善着色的真实性。为此,首先建立一种成像模型将光线颜色与物体本征颜色的作用解耦,并进一步构建一个在白色光线下像素色度-灰度关系模型。针对现有算法着色失真原因,提出基于色度修正的改进方法。考虑到每个颜色块的调整只需用户选择一个灰度参数即可实现,通过设置默认值自动实现粗调,并可通过人机实时交互进一步细调,避免了传统算法手工设置复杂层次颜色涂抹的问题。基于上述模型,提出一种实现有色光线下着色的简便方法。实验结果表明不仅着色图像经过修正后能很好的体现高光暗调区域的颜色特点,其着色质量甚至优于传统算法中利用手工设置复杂层次颜色涂抹所实现的光影效果,而且每次修正时间小于0.3s,使得整个调整过程可通过人机实时交互完成,说明该方法能提供一个提高图像真实性的有效平台。

Existing colorization algorithms can hardly provide pleasant visual effect in highlight or shadow areas. Motivated by the general imaging principle of natural color images, a realism improvement ap- proach based on post-process correction in chrominanee is proposed. First, an imaging model is built, where the influence of both light and objective on resulting image is decoupled. Then,a further chrominance-lurninance relation model for any pixel imaged in white light is established and employed to identify the very reason for falsehood resulting from existing colorization methods, and a realism improvement strategy is proposed based on chrominance coorection. Since only a grey parameter is to be set,two con- venient operation modes are available--automatic coarse adjustment based on default values, and fine tuning by user＇s interventioru In addition, this decupling model makes it easy to extend colorization to case of color light. Experiment results indicate that the visual effect after chrominance correction provides more accurate feature significantly in highlight or shadow areas and even outperforms that of existing al- gorithms under expensive and time-consuming miscellaneous input. The runtime of every correction is less than 0.3 s, which means this operation could perform as human-computer interaction in real-time as well, and hence demonstrates that this proposed method could provide a general platform to improve re- alism of colorization result.