This paper analyzes the performance of wavelet and curvelet transforms for image denoising against various types of additive noise, including Gaussian, Poisson, and speckle noise. The study reveals that curvelet transforms provide superior image quality, characterized by lower mean square error (MSE) and higher peak signal-to-noise ratio (PSNR) compared to wavelet transforms. Experimental results validate the effectiveness of curvelet transforms in preserving image details, particularly edges and linear features, during denoising.