Applying the FDR threshold filtering algorithm which is put forward by Felix Abramovich and Yoav Benjamini to image filter, on the basis of the above, for the image layered FDR threshold filtering algorithm is put forward in this paper. Through respectively making multiple hypotheses testing to the low frequency coefficients at the last level and the high frequency coefficients at other levels in horizontal, vertical and diagonal orientations, multiple threshold values are obtained At last, taking the gained threshold values as the criterion we could threshold all the coefficients at the individual level. The simulation results for images of lena, cameraman show that the proposed threshold algorithm is better than median filter, universal threshold, FDR and free distributed FDR threshold filtering algorithm in signal to noise ratio, peak signal to noise ratio and mean squared error .