Noise reduction is a critical step in image pre-processing. In this study, we employed a genetic algorithm (GA) to optimize the parameters of the Nonlocal Means Filter (NLMF), renowned for its ability to preserve edges, for the reduction of Gaussian noise. The optimization process utilized the peak signal-to-noise ratio (PSNR) as the fitness criterion and was tested on images corrupted with varying levels of Gaussian noise. The performance of the optimized filter was evaluated using key quality metrics, including mean squared error (MSE), PSNR, signal-to-noise ratio (SNR), and structural similarity index metric (SSIM). Results demonstrated that the optimized NLMF consistently outperformed traditional parameter configurations, achieving superior noise reduction while effectively preserving image edges. These findings highlight the significance of carefully selecting optimal parameters to enhance the efficacy of image denoising techniques.