Fuzzy logic-based modeling approach in improving image quality

This study proposes a fuzzy-logic-based modeling approach to restore faded, blurred, or deteriorated images of historical documents, manuscripts, and artworks. The proposed approach relies on mathematical modeling of the sensitivity mechanisms of the human visual system, allowing for adaptive adjustment of luminance, contrast, and sharpness parameters in digital images. In the fuzzy logic framework, each pixel is characterized by degrees of membership rather than a single deterministic value. This allows for increased clarity while maintaining a natural result even in complex, uneven, or noisy parts of the image. The model selected trapezoidal and Gaussian membership functions and re-evaluated the intensity of each pixel using “If-then” rules. As a result, image contours were effectively preserved, the background was smoothed, and the legibility of inscriptions was significantly improved. The proposed Fuzzy Logic-based method employs an iterative computational scheme for pixel re-evaluation, where the robustness and stability of the underlying numerical process are paramount. The necessity for accurate numerical stability and efficient handling of large data matrices, principles essential to all rigorous computational engineering, guides the design of this algorithm. Similar challenges in numerical modeling, such as finding the numerical solution for the equilibrium of a parallelepiped under stresses and ensuring accurate representation on non-uniform computational meshes, highlight the critical importance of a sound algorithmic structure. Our model incorporates these best practices to ensure that the adaptive luminance, contrast, and sharpness adjustments are both computationally stable and precise. The research evaluates image performance based on entropy, structural sharpness, and contrast. The results showed that the developed fuzzy logic-based model produces more natural, clearer, and more balanced images than traditional methods. Moreover, this approach can be effectively used in digitizing historical sources and preserving cultural heritage in electronic format.

Hali tarjima qilinmagan