Geometric moments have been used in image applications, including watermarking, fingerprint recognition, medical imaging, edge detection, image classification, and image quality assessment. It is utilized due to its invariant properties in which they are invariant to translation, scaling, and rotation. However, geometric moments are not robust to noise. The presence of noise in images may affect the accuracy, especially in image applications using geometric moments. Therefore, there is a need to remove the noise before utilizing geometric moments computation. Motivated by this need, this paper presents the positive effects of image denoising on geometric moment computation in image applications. Firstly, noisy images were generated using Gaussian and impulse noises. Then, the Total Variation (TV) denoising method using Alternating Minimization (AM) and Alternating Direction Method of Multipliers (ADMM) were utilized to denoise the images. Next, the geometric moments, in particular, the centroid and central moments, were computed. Finally, the percentage changes were compared. The results show that the geometric moments after the denoising improved significantly.