Kibern. vyčisl. teh., 2015, Issue 182, pp.
1Yuriy Fedkovych Chernivtsi National University (Chernovtsy)
2Bukovinian State Medical University (Chernovtsy)
MULTILEVEL METHOD OF LOCAL CONTRAST INCREASE AND IMAGES HETEROGENEOUS BACKGROUND REMOVAL
Introduction. The increase of local contrast and removal of heterogeneous background are the widespread problems of the digital image processing [1–4]. In existing local methods, such as the method of images adaptive contrast enhancement, a value of the local contrast is computed in vicinity of each pixel within a predetermined sliding window. The disadvantages of the existing local methods include poor performance, complicated selection of filter parameters and errors in the calculation of the intensity of the resulting image.
The purpose of the paper is to develop a multilevel method of local contrast increasing and removal of heterogeneous background of images with the high performance and accuracy using the minimal number of filter parameters.
Methods. The signal envelopes are calculated by linear and cubic approximation.
Results. The multilevel method of removing heterogeneous background and local contrast enhancement of images within the window of the Mw Ч Nw pixels size has been elaborated and developed in the MATLAB system . By means of the developed method the heterogeneous background has been successfully removed and local contrast has been increased for the test simulated and medical X-ray images.
Conclusions. Time of the image processing by the multilevel method is shortened comparing with per pixel processing in tR ~ (Mw Ч Nw)2 time, for example at the window size of Mw Ч Nw = 11 Ч 11 pixels the value of tR ≈ 10 times. The optimal distance between the windows centres on height and width is equal to SH0 = [Mw/2] + 1 and SW0 = [Nw/2] + 1, respectively.
Keywords: digital image processing, local contrast increasing, heterogeneous background removal.
- Gonzalez R., Woods R., Eddins S. Digital image processing. — M.: Technosphere, 2005. — 1072 p. (in Russian)
- Russ J.C. The image processing handbook. 6th ed. — CRC Press, 2011. — 817 p.
- Design features of medical information decision support system based on data mining / G.V. Knyshov, A.V. Rudenko, E.A. Nastenko & others // Cybernetics and Computer Engineering. — 2014. — Vol. 177. — P. 79–87. (in Russian)
- Bondina N.N., Muratov R.Yu. Adaptive filtering and image contrast changing algorithms // Vestnik NTU “KPI”, 2014. — №35. — P.35–42. (in Russian)
- Ketkov Y.L., Ketkov A.J., Schulz M. Matlab 7: programming, numerical methods. — SPb. : BHV-Petersburg, 2005. — 752 p. (in Russian)