Research on Application of Multispectral Color Copying Technology (I)

There are many theories for the evaluation of color reproduction. Among them, the famous British color expert Hunt, through the comprehensive analysis of various factors, the color reproduction is divided into six types: spectral color reproduction, color reproduction, correct color reproduction, equivalent color reproduction, corresponding color reproduction and Like color reproduction.

Although the conventional color copying method successfully realizes two methods of color reproduction and correct color reproduction, it is unable to realize the unconditional reproduction of color, that is, spectral color reproduction, because of the inherent characteristics of the metamerism. The multi-spectral color reproduction technique uses a multi-primary color imaging method to increase the color gamut by increasing the degree of freedom of color reproduction. The spectral matching reproduction method eliminates the problem of metamerism, thus achieving unconditional matching of colors. This technology is of great significance to the development of high-fidelity printing technology, and will certainly become the direction of future printing industry development.

First, the defects of the traditional four-color printing

Traditional printing uses CMYK four primary colors for color reproduction. Due to its limitations, four-color inks cannot cover all spectral information of the original color. In fact, traditional four-color printing is based on the principle of metamerism to achieve color reproduction.

The principle of metamerism refers to the fact that in the printing and copying, as long as the color of the printed document is the same as the color of the original, even if the spectral composition of the two is different, it can be copied and approved as the correct color. This method greatly reduces the difficulty of color reproduction, and in most cases can achieve the correct reproduction of colors, thus constituting the basis for the realization of traditional printing methods. However, when the illumination source and the observer change greatly, the copying effect tends to show a large deviation. This issue is also an important source of quality disputes in the traditional printing industry.

In addition, the color that can be reproduced by the four-color printing is only about half of the visible spectrum color gamut, and the effect of copying a large color image with a large dynamic range is often unsatisfactory. Under the current state of the art, even if people perform precise color management and color gamut matching during copying, the problems of metamerism and gamut too small cannot be solved fundamentally.

The traditional printing method focuses on matching the density of the original with the printed color density. However, this matching approach focuses too much on the discussion of the amount of ink and ignores the problem of matching the color and brightness of the copied color. In fact, due to the limitations of the color gamut of four-color printing, it is often in a paradoxical position when regulating color components and neutral gray components. It is inevitable to increase the color component ink to increase saturation. The neutral gray component is increased to reduce brightness. In other words, four-color printing tends to emphasize the reproduction of saturation excessively, and it is necessary to use the method of sacrificing the brightness.

Due to the above-mentioned defects of traditional copying, the effect of copying is often unsatisfactory, mainly due to the heavy color of the image, poor stereoscopic effect, more serious level loss, and different levels of color distortion.

Second, the advantages of multi-spectral color reproduction technology

The multi-spectral color reproduction technology achieves color reproduction by acquiring, analyzing, and processing multi-spectral data. This technology uses spectrum matching as a color reproduction standard, and enlarges the reproduction color gamut by increasing the degree of freedom of color superposition. Due to the uniqueness of the spectral reflection curve, the reproduction effect can be kept stable regardless of changes in the light source and observation conditions. In addition, the sampling of multiple bands of spectral reflectance can record color characteristics as detailed as possible, effectively solving the problem of too low data accuracy in the traditional mode.

In the field of color reproduction, the principle of color reproduction of digital devices such as monitors, printers, and scanners is very different from the formation mechanism of human eyes. Due to its own limitations, the traditional four-color printing model cannot solve the problem of color deviation fundamentally even with the aid of a color management system. The multi-spectral replication technology greatly improves the integrity of data acquisition by increasing the number of sampling channels, thereby achieving high-quality color reproduction.

In view of the above advantages, this technology has been successfully used in the field of copying and preservation of valuable art works as well as online shopping. At the same time, this technology also lays a solid foundation for future high-fidelity printing and cross-media publishing.

Third, multi-spectrum replication technology process

Multi-spectral color reproduction technology describes color information as spectral data by describing color spectral reflectance or transmittance. The specific process flow can be divided into the following steps:

1. Data acquisition

Use multispectral cameras with multicolor filters to capture multispectral image data of originals or objects. Typically, the acquisition system consists of a multispectral light source, a color filter, and a multispectral camera. Compared to the traditional three-color image acquisition method, this system has the following advantages:

The light source has a short start-up process, a wide spectrum, and a high radiation efficiency; the color filter has strong transmissivity and is free from background light interference; it can acquire high-resolution data, a variety of data support modes, and high imaging contrast.

After obtaining the spectral data, it needs to be analyzed and processed to achieve high-precision spectral reconstruction. The matrix representation of the mathematical model for multispectral data acquisition is now described as follows:

Let the spectral power distribution of the multispectral light source be S, and the spectral emissivity of the object be r, r = [r1, r2, ... rn]T, where n denotes the number of sampling wavelengths, and T denotes the transfer rank operation of the matrix. In a multi-spectral camera, the spectral transmission characteristics of the m color filters can be represented by the matrix F, and the spectral sensitivity of the detector is represented by the matrix D.

Integrating the above matrix, the colorimetric integral formula can be obtained, and the color value of the collected color is t=(DF)TSr. Subsequently, the color tristimulus values ​​XYZ and CIELAB can be obtained through corresponding linear and nonlinear transformations. Coordinates and other color values.

In addition to the methods described above, principal component analysis (PCA) can also be used to select the optimal filter design and achieve more accurate spectral reconstruction. This method is often used in the mutual conversion between the integral density and the analytical density in the photographic technique, and it is also commonly used in the establishment of a scanner's high-precision equipment property file. (To be continued)