This paper introduces a novel approach for automatically selecting the parameters of discrete Dual Hahn polynomials using a swarm optimization algorithm. The proposed method identifies the optimal parameters for constructing Dual Hahn moments, thereby improving the accuracy of image moments. Our approach minimizes the mean square reconstruction error, leading to significant improvements in image and signal reconstruction quality. The results demonstrate that the proposed method achieves lower reconstruction errors, especially at lower orders of dual Hahn moments. Additionally, the optimized dual Hahn moments were evaluated and compared against other parametric discrete moments, both with and without parameter optimization. The experiments clearly highlight the advantages of this approach, particularly at lower moment orders.