Enhancing Visibility in Foggy Landmark Images: Advanced Restoration Methods

The proposed solution to improve the clarity and detail of foggy or hazy images that show historical places, monuments, and archaeological digs an important task for records and preservation. Initially, the effectiveness of two distinct algorithms is explored: Markov Random Fields for image restoration and a Riemannian rank-adaptive method for low-rank matrix completion. Through separate implementations, their individual capabilities in restoring image clarity and detail are evaluated. Afterwards, both algorithms are integrated to synergistically enhance the restoration process. The findings of our proposed solution demonstrate that the combined approach yields significantly improved results, effectively mitigating haze and revealing intricate details obscured in the original images. The integrated methodology offers a promising solution for the restoration and preservation of invaluable cultural heritage imagery. The Combined Restoration approach has the lowest entropy of 6.0 for the enhanced image, indicating a more structured result. It achieves the highest visibility metric of 29.12 and a competitive MSE of 109.93 compared to the hazy image. The proposed system performs best in terms of balancing entropy and visibility enhancement while improving the difference (MSE) between the hazy and restored images.