An improvement to the efficiency of the nonlocal means (NLM) denoising method is proposed in this paper. The NLM denoising method replaces each pixel by the weighted average of pixels with the surrounding neighborhoods. We employ a cosine weighting function instead of the original exponential function used in the weight calculation. Experiments comparing the proposed function against the original function to demonstrate its superior denoising performance in terms of peak signal to noise ratio (PSNR) and histogram, using various test images corrupted by additive white gaussian noise (AWGN).