Measuring the microstructure surfaces of mechanical, optical, and opto-electromechanical details contributes to surveying, evaluating, and improving the quality of manufacturing, assembly, and operation of systems. The use of the white light interference (WLI) method in measuring microstructure surfaces of optomechanical details has many meanings in practical applications and will be focused on the development by improving processing algorithms and signal analysis methods, alternating with different methods. Both improving measurement speed and increasing measurement accuracy are the important factors that are being focused on development. Processing interference fringe signals to eliminate the influence of the noise from the environment and therefore, the measurement accuracy increases. In this paper, we propose an interference fringe signal processing method that significantly eliminates the effect of noise. The firstly, the signal is converted to the frequency domain by using Fourier transform. Next, the windows are used to remove unnecessary noise signals, but the useful signal at the three lobes will be saved. This signal is again converted by using the inverse Fourier transform. Finally, the envelope method will be used to find accurate information about the height of the surface profile. The simulation results for different surface profiles are shown. The results achieved in terms of accuracy of the proposed method are 1.8641 nm, which is forty-five times smaller than the traditional fast Fourier transform (FFT) method of 84.9641 nm at a high noise level.