节点文献

基于高光谱图像处理的中国书画鉴赏试验研究——以齐白石《牡丹》为例

Experimental study on the appreciation of Chinese calligraphy and paintings based on hyperspectral image processing: taking Qi Baishi’s Peony as an example

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 唐兴佳胡陈宇帆高婧威董文强杨文宗

【Author】 TANG Xingjia;HU Chenyufan;GAO Jingwei;DONG Wenqiang;YANG Wenzong;Key Laboratory of Archaeological Exploration and Cultural Heritage Conservation Technology(Northwestern Polytechnical University) ,Ministry of Education;NPU Institute of Culture and Heritage (NICH);School of Aeronautics,Northwestern Polytechnical University;School of Materials Science and Technology,Shaanxi University of Science and Technology;

【通讯作者】 胡陈宇帆;

【机构】 考古探测与文物保护技术教育部重点实验室(西北工业大学)西北工业大学文化遗产研究院西北工业大学航空学院陕西科技大学材料科学与工程学院

【摘要】 《牡丹》是齐白石花鸟画的典型代表。该画作所绘牡丹墨彩鲜艳而饱满,红色、黑色等多色搭配,形成了强烈的视觉冲击力和独特的“红花墨叶”画风,更通过墨色的浓淡变化与多层运用,使得牡丹花瓣与枝叶更加富有层次感和深度。本研究利用高光谱成像技术真实、全面、多维度采集该书画所承载的视觉与材质信息,并借助高光谱图像处理技术分析该画作的绘制工艺与视觉表达特点。通过书画高光谱图像的典型书画元素间的单通道图像和高光谱曲线对比,分析书画绘制材料和工艺的差异;利用高光谱图像不同波长灰度图像和伪彩色合成图像,以及高光谱主成分变换灰度图像和伪彩色图像,增强并分辨书画中一些人眼不可见信息,进而研究其绘制工艺特点;利用高光谱识别技术,在无颜料标准光谱数据库时,利用画作自有颜料光谱,进行同类别颜料的分布区域识别,进而为后续可能的书画褪色修复和长期健康状态监测提供依据,并为必要情况下的书画鉴别提供判据;利用高光谱图像监督分类,更好地了解书画不同颜料、不同图案的多重分布和整体布局;利用“色度-饱和度-亮度”(hue saturation intensity, HSI)变换的灰度图像及其伪彩色图像发现书画元素绘制的渐变分布方式,以及多层叠加关系。研究结果表明:《牡丹》画作中,齐白石运用墨彩的深浅、浓淡、轻重、干湿和线条的虚(渲染)实(勾线)结合以及多重绘制等表现牡丹的质感和空间感,且画作中采用了不同印泥材质的印章,并在纸张上表现出组合式特点。上述研究方法与研究结果为深入研究齐白石绘画特点、绘制工艺及艺术表达等问题提供了新的技术方法与分析案例。

【Abstract】 Qi Baishi’s Peony is a representative work among his flower and bird paintings. The bright and vivid tones of the peony, with the use of multiple colors such as red and black, create a strong visual impact and embody his distinctive “red flower and ink leaf” painting style. The variations in ink intensity and the multi-layered application of color lend to the peony petals and branches a sense of depth and texture. In this study, hyperspectral imaging(HSI) technology was applied to comprehensively capture the visual and material information contained in Peony in a real, comprehensive, and multi-dimensional manner. Using HSI processing, the painting process and visual expression characteristics of the painting were analyzed. By comparing single-channel images and hyperspectral curves of representative calligraphy and painting elements, the differences in materials and techniques used in the painting were identified. Grayscale images and pseudocolor composite images at different wavelengths, along with hyperspectral principal component transformed images, were used to enhance and distinguish previously invisible information, providing insight into the artist’s painting process. When no standard spectral database of pigments was available, hyperspectral recognition was employed to identify pigment distribution areas by extracting self-referenced spectral features directly from the painting. This provides a scientific foundation for future conservation efforts, including color fading restoration and long-term condition monitoring, as well as for authentication studies. Supervised classification based on hyperspectral images data enabled a clearer understanding of pigment distributions and compositional layouts. Furthermore, grayscale and pseudocolor HSI transformations revealed gradient distributions and multi-layered stacking relationships among pictorial elements. The results demonstrate that in Peony, Qi Baishi skillfully integrated variations in ink tone—depth, concentration, lightness, dryness, and wetness—with the dynamic interplay between “void”(rendered) and “solid”(outlined) strokes. The painting also exhibits multiple layering techniques and the use of seals made from different ink materials, reflecting a hybrid stylistic approach on paper. Overall, this study offers new technical tools and analytical perspectives for in-depth exploration of Qi Baishi’s artistic techniques and expressive methods.

【基金】 国家自然科学基金面上项目(52372035);国家重点研发计划项目(2023YFF0906202)资助
  • 【文献出处】 文物保护与考古科学 ,Sciences of Conservation and Archaeology , 编辑部邮箱 ,2025年06期
  • 【分类号】TP391.41;J212
  • 【下载频次】19
节点文献中: 

本文链接的文献网络图示:

本文的引文网络