TV-Regularized Sparse Reconstruction Minimization Algorithm Based on ADMM

The ADMM-based Total Variation (TV) regularized sparse reconstruction minimization algorithm is an optimization method widely used in signal processing and image reconstruction. This algorithm combines the edge-preserving characteristics of Total Variation with the efficient solving capability of the Alternating Direction Method of Multipliers (ADMM), making it suitable for recovering high-quality signals or images from sparse or noisy observations.

In sparse reconstruction problems, TV regularization maintains structural smoothness by constraining the gradient sparsity of signals while avoiding excessive smoothing of important features like edges. The ADMM framework decomposes the original problem into multiple subproblems, alternately optimizing primal variables, dual variables, and Lagrange multipliers, thereby simplifying complex high-dimensional optimization tasks. In implementation, the algorithm typically involves iteratively solving a TV-regularized minimization problem using proximal operators and updating dual variables with a step-size parameter.

The main advantages of this algorithm lie in its flexibility and convergence properties, enabling it to handle non-smooth objective functions and scale to large problems. In fields such as medical imaging, compressed sensing, and computer vision, the TV-ADMM method has proven effective in enhancing reconstruction quality while reducing artifacts and noise impacts. Key implementation considerations include proper parameter tuning for the penalty term and convergence criteria checks using relative residual norms.