No abstract

The article develops the process of modeling color separation in the reproduction of graphic information in printing technology, shows the management of the technical system at each stage. With the use of information technology, this complex and multi-stage technological process of color conversion is carried out automatically, which does not allow showing the clarity of technological operations.

Protection of goods and packaging from counterfeiting and copying, tracking their movement requires improvement of existing labeling and security methods and the development of new ones. Making changes to the image at the prepress stage is the cheapest and easiest way of protection compared to using special printing techniques, special substrates, and inks or additional tags such as RFID. In the article, we suggest a new method to create security printed features, identify them in prints, and confirm the authenticity of the image. The method uses a combination of regular (AM) and stochastic (FM) screening in one image. There are two ways of separating images for AM and FM screening. First is to choose several random intervals in shadows of image tonal distribution and in accordance with values in these intervals original image is separated into two. The second is to separate by structure, for example, use FM screening on edges or textures. We tried Canny edge detector and local binary patterns. By using random values as the parameters, it is possible to generate unique print runs or even individual prints using digital printing. And large variability in the areas of separation gives reason to consider that the suggested method is reliable. Fourier analysis in the suggested method allows not only to detect the presence of security printed features but also to confirm the authenticity of the image on a print. Authentication is implemented by obtaining a digital image of the print by scanning or photographing and comparing the spectral composition of the original image and the digital image of the print. An expert survey showed that after our method presence of a combination of AM and FM screening in images on prints is barely visible. As a result, this method can be used to protect packaging labels with images from copying.

A method of image preparation for printing reproduction is suggested. This method allows to automatically compensate transformations that occur during reproduction, by analyzing a histogram of test chart image and based on it, creating a compensation pre-correction function. It also takes into consideration the visual perception of images. Pre-correction function is applied to images at the prepress stage after all other corrections. It is aimed to compensate defects, occurring at the printing stage, caused by the process of tone value increase and restriction of tonal range reproduction. It is suggested to use a test chart, which is a gradient with an even increase of lightness in the range from 0 to 255. After printing the test chart its digital image is created by scanning. Then Gaussian filter is applied to the image with parameters according to the visual perception, and lightness distribution histogram is calculated. This histogram will have changes in lightness distribution in comparison with the original digital image. These changes will correspond to the influence of tone value increasing process during printing. The cumulative sum is calculated from the received histogram, and the pre-correction is being formed. And this precorrection applies to an image, prepared for printing in similar conditions as test chart. The algorithm was written on Python and allows to create a pre-correction using a press sheet with the test chart. It is shown that the use of the suggested method gives a positive result and doesn’t require expensive measurement equipment. Having a scanner calibrated for linear transmission of lightness and developed programming module is enough. This method was tested on electrographic printing equipment on three different types of paper. Statistic parameters of a histogram, such as mean, standard deviation and the Skewness, were used for evaluation. It is shown that the suggested method can be used as part of an automatized system based on histogram methods for image preparation before printing.