Doppler Asymmetric Spatial Heterodyne(DASH) spectroscopy is a new hyperspectral remote detection technology, which requires the data processing technology should also have corresponding super high precision. From the viewpoint of digital signal processing, an adaptive frequency tracking method was proposed, which is based on the spatial frequency of the signal to compensate the phase shift of the fringe signal, and recursive iterate to the phase information most close to the real value. Verified by simulation experiment contrast, the results show that, under the low intensity noise condition, the frequency tracking algorithm can improve the accuracy of signal frequency and phase extraction by about more than 100 times to compare with the traditional Fourier transform method, which proves it can effectively reduce the system error of DASH spectroscopy.