Research of Image Processing Methods in Publishing Output Systems
For automatic processing of images in publishing output systems it is necessary to pre-reduce the noise level, to increase sharpness, to exclude the possibility of a moire, that is, to improve the quality of the image before printing. There are the results of using the Fourier transform for simple figure images. The amplitude spectrum values for blurry and detailed images are considered, and the methods of automatic processing of images in laser-type publishing systems are also shown.
The frequency parameters that significantly impair the quality of the image include noise. It is necessary to estimate the noise level in the image and its type (analogue/ pulsed), because noise reduction operations depend on their type. In many cases, it is necessary to take into account the presence of deterministic noise caused, for example, by the raster structure of printing reproduction, if the original for obtaining a digital image is an imprint. Originals scanning should be made with high resolution to prevent the occurrence of a moire. The same problem can also be caused by periodic structure of the image itself. To solve this problem it is necessary: 1) to use filters to remove the raster structure; 2) to try saveing an existing raster structure for use on output.
The correction of the structural properties of the image is divided into: 1) sharpening correction; 2) noise correction. In most cases, the results for analysis, which are important for practice, can be obtained for a binary image, which is formed by the original multi-color or color image. This significantly simplifies image analysis. The amount of information processed is reduced and at the current level of computing development it is possible to work in real time. To improve the quality of rendering or transforming images to a comfortable appearance (for further processing in the spatial area) in laser output systems, it is possible to use methods for processing images that can be implemented in the area of spatial frequencies. The image is described by the matrix of the brightness values of the discrete elements in the spatial domain, then its representation in the spatial frequency domain will be a matrix, obtained by decomposing the output matrix of an image on a chosen basis. The periodic spectrum obtained after sampling is the spectrum of a digitized image. Next there is used the image processing method of decomposing the images into a number of separate harmonic components using the Fourier transform .In the case of images representing a discrete two-dimensional signal, the spectrum is also a discrete two-dimensional signal.
In order to form a high-quality, convenient image analysis, it is necessary to determine in advance which details of the image are perceived particularly clearly, and which are secondary to the analysis. Then, the comparative analysis of the spectra of the images of the observed objects and the background structure allows to use the frequency domain for a filtering operation. It is important for removing noise in laser-type publishing systems, as well as moire with a given loss of image detail.
 M. Ryzhykov. “Formation and obrabotku Pictires in laser systems Vision”. Students book. – SPb: NUAP, 2013. – 210 с. — SPb., 2013.
 O. Tymchenko., B. Havrysh. “Analysis methods for forming halftone dots in prepress systems publications”. Computer technology of printing. Collected Works. — Edition 2 (34). – Lwiw: UAP. – 2015. – 232 с. – С.89-96.
 M. H. Yang, D. J. Kriegman, N. Ahuja. “Detecting faces in images”. IEEE Trans. Pattern Analysis and Machine Intelligence. — 2002. — Vol. 24. — No. 1. — P. 34–58
 C. Q. Petriu, X. Yang. “A Comparative Study of Fourier Descriptors and Hu’s Seven Moment Invariants for Image Recognition”. CCECE. — 2004. — P. 0103–0106.
 Thawar Arif, Zyad Shaaban, Lala Krekor, Sami Baba. “Thawar Arif Object classification via geometrical, zernike and legendre moments”. Journal of Theoretical and Applied Information Technology. — 2009. — Vol. 7; No. 1. — P. 31–37.
 A. Nobuyuki, S. Shunji and T. Tadamasa, “Position annotated image reporting system for electronic textbook of illustration and image altus,” Proceedings 5th International Workshop on Enterprise Networking and Computing in Healthcare Industry (HealthCom), 2003, pp. 133-136.
 L. Mandic, S. Grgic and I. Srdic, “Data formats in digital prepress technology,” International Symposium on VIPromCom Video/Image Processing and Multimedia Communications, 2002, pp. 437-440.
 P. E. Wellstead and G. R. Wagner, “Input/output self-tuning algorithms applied to image processing,” 26th IEEE Conference on Decision and Control, Los Angeles, California, USA, 2007, pp. 1204-1207.