Abstract: As the main prevailing denoising method, how the threshold function works and whats the threshold value are the greatest importance techniques. Consider the dependencies between the coefficients and their parents, a non-Gaussian bivariate distribution is given, and corresponding nonlinear threshold function is derived from the model using Bayesian estimation theory. According to non-subsampled Contourlet transform and bivariate threshold function, a novel Non-Subsampled Contourlet Transform based on Bivariate threshold function (NSCTBI) for image denoising is proposed. This scheme achieves enhanced estimation results for images that are corrupted with additive Gaussian noise over a wide range of noise variance. To evaluate the performance of the proposed algorithms, the results are compared with existent algorithms, such as non-subsampled Contourlet transform and wavelet-based bivariate threshold function method for image denoising. The simulation results indicate that the proposed method outperforms the others 0.5~2.3dB in PSNR, and keep better visual result in edges information reservation as well.