A new uniform threshold function of wave-shrink for image denoising is proposed. The new function has the advantage of satisfying the shrinkage condition. By changing the parameter u,we can change the trend of this new function, which makes it possible to find an adaptive method whenever using the wavelet shrinkage method for image denoising. Finally, efficient formulas for computing mean , variance and risk of the new uniform threshold function are derived. Numerical experiments and comparative studies are performed, demonstrating the effectiveness of our threshold function. Additionally, the proposed new function can also give better performance in terms of both PSNR and in visual quality.