A Comparative Study Of Chinese Painting Color Reconstruction Based On BRDF And KECA Spectral Reconstruction Algorithm

As the only civilization among four major ancient civilizations,our country still exists and flourishes,in her various kinds of art inherited to the present through her long history,Chinese painting is extremely unique and valuable.As early as two thousand years ago,the prototype of Chinese painting had begun to appear,among all the extant paintings of traditional Chinese painting,ancient authentic works are incomparable in the value of art and history.However,due to the nature of the material itself,the older the Chinese painting works become,the worse the degree of perfection is.In the process of researching and displaying traditional Chinese paintings,the damage to Chinese painting is accelerated and irreversible with current preservation techniques,the use of modern technology to digitally copy it will greatly protect the authentic Chinese painting away from oxidation,moisture,friction and other natural or man-made damage,therefore,it is important to research the technology.People can see the colorful world because the surface of the object reflects wavelength from the light source,the pyramidal cell of human eyes receive the reflected wavelength and feed it back to the brain,then people have a perception of color.Objective factors that affect color perception include the light source,the reflective properties of the surface of the object,and the structure of the human eyes.In the process of color science technology development,people found the phenomenon of metamerism.If we want to cross-media reconstruct the real color,it is necessary to reduce metamerism as much as possible of the spectrums.The reconstruction requires the study of the surface reflection characteristics of the object to get the color information including spectral informationIn the current field of cross-media color reconstruction,using BRDF for color reconstruction is a hot technology,because BRDF can reflect the texture,material and spectral information of the surface of the object.After decades of research by scholars in various countries,several models have been developed to adapt to different situations since the birth of BRDF,so BRDF is very promising.This paper illustratesthe principle of BRDF reconstruction color through theory,then set up a BRDF information collection system to measure the relevant information then reconstruct the sample images of Chinese painting and color information.Multi-spectral imaging is a "image and spectrum integration" technology,it has been widely used in cross-media color reconstruction in recent years.Spectral reflectance reconstruction is a core issue in the field of multispectral imaging,a variety of algorithms have been used by scholars to continuously improve the effect and efficiency of color reproduction around this topic.In these algorithms,the kernel entropy component analysis algorithm has become the focus of colorists as a novel and efficient nonlinear processing algorithm.This paper introduces a spectral reflectance reconstruction method based on kernel entropy component analysis algorithm,a multispectral imaging system was built and multispectral data was acquired,then the sample image of traditional Chinese painting was reconstructed by using multi-spectral data and kernel entropy component analysis algorithm.In this paper,I build BRDF and multi-spectral imaging system respectively,measured and processed the data of the Chinese painting sample,then got the corresponding Chinese painting color block reconstructed image.The experimental methods were compared,and then the subjective evaluation and objective evaluation of color reconstruction effect were conducted.The experimental and analysis results show that the color images reconstructed by the two color reconstruction methods are all in good condition and are all within the range of acceptable color difference of the cross-media color reconstruction human eyes.In the experimental operation,BRDF-based color reconstruction method is simpler;in the color reconstruction accuracy,the image reconstructed by KECA algorithm is more sophisticated.