S.A. Shvidchenko, D.A. Bezuglov

A new algorithm of marking fragments contour on the occasional background with the use of mathematic apparatus discreet wavelet transformation has been worked out. Mathematic modeling is carried out and suggested algorithms are tested on the series of test images. The analysis of the results of mathematic modeling gives opportunity to come to the following conclusion: the suggested method of image processing can effectively allocate image contours distorted by noise. A new method and algorithms of wavelet image differentiation in background noise using a discreet wavelet transformation can improve the ratio of the peak signal - noise from 4,8 to 9,6 Db to 3-4 Db and reduce the standard deviation error. In this case the wavelet transformation eliminates the use of various masks that is, in fact, abandon the ineffective method of numerical differentiation. On the basis of this method other contour marking algorithms based on the wavelet differentiation with other wavelet bases can be implemented. Thus, the scientific task of new methods of automated analysis of the results measurement development for marking contour fragments in images having noise is solved. The proposed methods and algorithms can be used in developing the systems of signal processing in digital images and automated robots in industry in conditions of observation, which complicates the registration process when there is no apriori information of background noise.

 

1. Bezuglov D.A., SHvidchenko S.A. Informacionnaja tekhnologija vejjvlet-differencirovanija rezultatov izmerenijj na fone shuma // Vestnik kompjuternykh i informacionnykh tekhnologijj. 2011. № 6(84). S.42-45.
2. Bezuglov D.A., Rytikov S.JU., SHvidchenko S.A. Metod vejj­vlet-differencirovanija v zadache vydelenija konturov // Sovremennye problemy radioehlektroniki. 2012. № 6. S. 52-57.
3. Bezuglov D.A., Rytikov S.JU., SHvidchenko S.A., Gavrin M.S., Gavrin D.S. Vydelenie konturov izobrazhenijj v informacionnykh i upravljajushhikh sistemakh s ispolzovaniem metoda vejjvlet-preobrazovanija // Nelinejjnyjj mir. 2012. № 11.S.846-852.
4. Bezuglov D.A., Rytikov S.JU., Cugurjan N.O., SHvidchenko S.A. Svidetelstvo № 2010613246 o gosudarstvennojj registracii programmy dlja EHVM ot 17.05.2010 «Differencirovanie signalov vejjvletom DOG». Zajavka №2010611469 ot 17.05.2010.
5. Bezuglov D.A., Rytikov S.JU., Cugurjan N.O., SHvidchenko S.A. Svidetelstvo № 2010613245 o gosudarstvennojj registracii programmy dlja EHVM ot 17.05.2010 «Differencirovanie signalov vejjvletom WAVE». Zajavka № 2010611468 ot 24.03.2010.
6. Bezuglov D.A., Andrjushhenko I.V., SHvidchenko S.A. Svidetelstvo № 2010613399 o gosudarstvennojj registracii programmy dlja EHVM ot 21.05.2010 «Algoritm optimizacii chisla koehfficientov vejjvlet-preobrazovanija». Zajavka № 2010611409 ot 22.03.2010.