Besides the regular biological or biomedical imaging applications, time-gated fluorescence lifetime imaging microscopy (FLIM) could be used to image living cells or rapid biochemistry reactions by its advantage of wide field and obvious parallel acquisition speed. As the basis of system detection mode, a good fitting algorithm could satisfy the lifetime precision and acquisition demand at the same time. The common used fluorescence lifetime fitting methods include rapid fluorescence lifetime determination(RLD), linear least square method (LLS)and iterative nonlinear least square (NLLS)method. In this study, we compared these three methods and analyzed the lifetime results. It demonstrated the RLD algorithm could achieve a fast processing speed and a good contrast of lifetime map, however, it shows the results of the rapid fitted lifetime map exists some distributed noise caused by insufficient time-resolved fluorescence intensity images. To remove these noise spots and preserve the lifetime map precision, we chose a four-order partial differential equations (PDEs), which achieved a trade-off between noise-removing and lifetime map feature. In this paper, we proposed a rapid fluorescence lifetime acquisition method combining rapid lifetime determination and four-order PDEs algorithm based on time-gated FLIM, which could satisfy the high speed detection and lifetime precision demand of living cells or rapid biochemical reactions.