The Department of Mathematics at the College of Education for Pure Sciences organized a seminar titled "Application of the Convex Regularization Model in Image Processing." The seminar, presented by postgraduate student Ashraf Yassin Jumaa, focused on the study and development of an advanced mathematical model for digital image restoration—specifically for removing defects and distortions. The model is based on the Mumford–Shah framework, utilizing the Ambrosio–Tortorelli variational formulation—an approximation method used to address variational problems related to edge detection and the preservation of an image's geometric structure during the restoration process.
The research aims to evaluate the efficiency of the proposed model by applying it to digital images across eight different color spaces. It includes comparisons between processing in the spatial domain and the frequency domain, while examining how the choice of color space and processing domain influences execution speed, the accuracy of the results, and the quality of the restored image.
From a mathematical and numerical perspective, the work employs the finite difference method to approximate the equations associated with the model, transforming them into numerical systems that can be solved computationally. Furthermore, the study addresses the fundamental theoretical aspects of the numerical system—including stability analysis, the existence and uniqueness of solutions, and the convergence of the numerical scheme—thereby providing a rigorous mathematical foundation for assessing the reliability and accuracy of the resulting solutions. The numerical results, within the experimental cases studied, demonstrate that processing in the frequency domain achieves higher computational efficiency in terms of execution speed. Furthermore, it yields improvements in several image restoration quality metrics—specifically, a reduction in the Mean Squared Error (MSE), alongside higher Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM) values. The results also indicate the model’s ability to preserve edges and significant geometric details—a crucial aspect for image restoration applications that require the recovery of lost information while minimizing the distortion of original structures and patterns.
The significance of this work lies in its integration of variational formulation, mathematical analysis, numerical methods, and experimental evaluation, incorporating a comparative study across various color spaces and processing domains. This approach contributes to a deeper understanding of how color representation and numerical processing choices influence the efficiency of image restoration algorithms, and it supports the development of more accurate and efficient methods for processing degraded images in scientific and engineering applications

.