my country is a country with an ancient civilization of 5,000 years. In the long history, people have preserved these cultural civilizations in a large number of calligraphy and paintings, and murals, and showed its beauty and value to the present and future generations. In these rich calligraphy and paintings, colorful colors are the materials that people give themselves to imagine and create boldly. This colorful space is vividly expressed by using different pigments and different techniques. If the painting theme is the spiritual source of each calligraphy and painting and mural, then the rich pigments are the strong material support of each calligraphy and painting and mural. The study and identification of ancient pigments is a process of studying the development of ancient civilization. The use of pigments in different dynasties and regions is closely related and constantly moving forward. The age of cultural relics can be inferred from the material identification of pigments, and the change of dynasties and regional development can be inferred from the use and changes of pigments between different cultural relics. This paper uses hyperspectral imaging technology to conduct non-destructive identification research on ancient pigments, uses hyperspectral cameras to collect calligraphy and painting and mural data, and non-destructively identifies pigments by building a spectral database and spectral matching technology.

This study applied a 400-1000nm hyperspectral camera, and the FS13 product of Hangzhou Caipu Technology Co., Ltd. can be used for related research. The spectral range is 400-1000nm, the wavelength resolution is better than 2.5nm, and up to 1200 spectral channels. The acquisition speed can reach 128FPS in the full spectrum, and the highest after band selection is 3300Hz (supporting multi-region band selection).

The feasibility of identifying cultural relics based on pigment reflectance spectroscopy should start from the microscopic structure of the pigment layer of murals and calligraphy and painting. The pigment layer of murals and calligraphy and painting is composed of particles (pigments), adhesives and additives. The role of adhesives is to glue the particles together and improve the mechanical strength of the material. Additives are used to improve the performance and working properties of the material.

As cultural relics have been baptized by the years and exposed to the air for a long time, the binder in the pigment layer has gradually deteriorated and oxidized. Therefore, the data collection of the pigment layer of cultural relics can be roughly attributed to the collection of particulate matter (pigment) alone. Therefore, the identification of pigments based on reflectance spectroscopy naturally eliminates the influence of binders on the pigment spectrum, so that the spectra of cultural relics pigments can be compared by establishing a standard pigment spectrum library from the laboratory to obtain the identification results. A more complicated but absolutely effective way to identify pigments using spectral reflectance is to establish a large spectral database. The spectral database should not only have spectral data of pure pigments, but also spectral data of mixed pigments of different kinds. At the same time, the influence of glue and alum on the pigment layer should be considered. Therefore, a large amount of experimental data should be collected, but this alone is far from enough, because the pigment layer on the surface of ancient cultural relics is complex, not only pigments, but also oxidized glue. In the long-disrepaired temple murals and underground tomb murals, there will be other dust in the pigment layer, and the remnants of bacterial corrosion, which are extremely common on the surface of cultural relics. Due to the complexity and variability of these multiple factors, it is not a one-day job to build a complete pigment spectrum library, but it requires years of supplementation and improvement. Fortunately, when I was studying a Ming Dynasty calligraphy and painting, I found that the spectral matching results of the relatively pure mineral pigments on the calligraphy and painting were very good, so the idea and focus of this study is to study the spectrum of pure pigments. The partial map of this Ming Dynasty calligraphy and painting and the spectral matching results are listed as follows:

The data collected by the hyperspectral camera is also called a hyperspectral image cube. It is composed of images of different bands. The direction of the arrow is its band direction, from the visible light band to the near-infrared band. The data is composed of dozens or thousands of images of different bands in the same area superimposed together. The spectral curve of any pigment can be obtained through the hyperspectral image data, and then the distribution of the target substance in space can be obtained through the spectral curve of any pigment using the classification algorithm. We select pixels at different pigments in the image to obtain the spectral curve corresponding to the pixel. We select the spectral curve corresponding to gold, and then use the spectral angle filling algorithm to set the matching similarity threshold to 0.1 to obtain the distribution of gold pigment in the image space.

Effects of instrument noise

In the process of data collection of actual instruments, since a lot of data is collected, it is impossible to capture a corresponding dark current file for each image. Studying the changing law of instrument noise helps to reasonably allocate the time for collecting dark current data during the project collection process, which saves collection time and improves work efficiency while ensuring the quality of data collection.

The experimental plan is: place the camera indoors, simulate the outdoor collection environment to capture dark current data, and strictly cover the camera lens with the lens cap. Collect dark current data once every interval. The dark current time collection setting plan is as follows:

The collected dark current data is processed as follows: the dark current at other times is subtracted from the first dark current, and the average value of each band (400nm-1000nm) of the subtracted data is calculated, and then the average value of each band is plotted into a curve to observe the changes in dark current in different bands. Experimental results:

The experimental results show that the average value of the dark current data in each band (400nm-1000nm) after subtraction is almost 0, and the maximum

fluctuation does not exceed 0.4, which means that the change of dark current in different time periods is almost zero, and it has almost no direct impact on

the change of experimental data.

Experimental inspiration: The working performance of the hyperspectral camera VNIR400H is stable, and the dark current does not change much. It remains

at a stable value. There is no need to consider the influence of the noise data of the instrument too much. You can shoot 1~2 times in a short interval (one

acquisition area).

Summary This chapter controls and experiments the parameters that affect the data of the hyperspectral camera in order to obtain accurate image data and

spectral data. Through experiments, it is found that the influence of illumination and the dark current of the instrument on the data is not great, and the focal

length has almost no effect on the reflectance spectrum of the corrected data pixel, while the exposure time will have a certain influence on the formation of

spectral data. Through a series of experiments, we have a better understanding of the performance of the instrument, which also provides a strong guarantee

for subsequent data collection and the establishment of a pigment spectrum library.