ANALYSIS OF WAVELET TRANSFORMATION METHODS IN IMAGE COMPRESSION PROBLEMS

In this work, an analysis is carried out using spectral transformations to obtain compression indicators and signal-to-noise ratio, the best wavelet basis, namely, a special case of the Fourier transform for image compression according to the signal-to-noise ratio criteria. The increase in the compression ratio with increasing Dobeshi order is shown due to the fact that increasing the order increases the scaling function, which allows to increase the degree of compression of the image, obtaining a satisfactory quality of this image. But with increasing scaling function, the length of the filter increases, which complicates the implementation of this method. Spectral transformations in the problems of image compression in modern algorithms are shown that they can increase the compression ratio of black and white and color images with a comparative visual quality in relation to the algorithms of the previous generation, based on discrete cosine transform. Also, the design of a mesh volumetric object in two-dimensional coordinates is carried out to remove invisible vertices and segments. A study of the transfer of the remainder in the two-dimensional field of vertices during alternation and sequential television scans. It is shown that in order to reduce the data flow, it is advisable to perform a spectral wavelet transform before transforming a three-dimensional grid image into a two-dimensional one. By removing insignificant values of the wavelet coefficients, it is possible to achieve compression by a factor of 5, while the image quality represented by the signal-to-noise ratio reaches 35 dB – an estimate of the indicator of acceptable visual quality for comfortable viewing.