A adaptive thresholding algotithm for motion detection based on two-dimensional spatio-temporal fuzzy entropy principle is proposed.A fast solution for calculating the two-dimensional fuzzy entropy threshold is deduced,in which repeated calculation is avoided by using integral and iterative operation,and the various calculation operation is reduced from O(N⁴) to less than O(N³) for N gray-level image.The motion detection is reduced to two binary partition problems,and two-dimensional spatio-temporal fuzzy entropy principle is used to determine the threshold T,where the explicit function form or parameters of background distribution are not needed to be known.The experimental results show that the information of moving objects which have low contrast to background can also be extracted completely by this method in real time.