Fuzzy Logic Approach to Improving the Digital Images Contrast
Denys Mikhov, Yuriy Panteliyovych Kondratenko, Galyna Kondratenko, Ievgen V. Sidenko · 2019 IEEE 2nd Ukraine Conference on Electrical and Computer Engineering (UKRCON) · 2019
In this paper, methods of increasing the contrast of digital images based on fuzzy logic as soft computing technique are considered. The existing classical methods of increasing the contrast, such as the minimum-maximum linear contraction stretching, the general and adaptive methods of equalization of the histogram, the methods for increasing the contrast of the fuzzy sets “white pixel” and “gray pixel” are analyzed. For comparing analysis all methods were evaluated using metrics such as contrast per pixel, Lab Variance, and peak signal-to-noise ratio. The analysis of these methods revealed their advantages and disadvantages: (a) equalization methods greatly increase the contrast, but deform the color component, and also add artifacts to images, in turn, (b) methods based on fuzzy logic less increase contrast, but lack disadvantages in the deformation of the color component. In this paper, a combination of “white pixel” and Gamma-Correction methods is proposed, that leads the brightness of the image to the average, and then increases the contrast of the image based on fuzzy logic technique.