Abstract: As a high-resolution spectral analysis technology, spatial heterodyne spectroscopy is widely used in atmospheric detection, satellite remote sensing and related fields because of its ability to obtain ultra-high spectral resolution data. To reduce the blind data in the interference signal obtained by spatial heterodyne spectrometer and improve the accuracy of subsequent spectral reconstruction, a correction method combining sliding-threshold defection and column-window median filtering is proposed. Compared with the principal component analysis algorithm and the average substitution algorithm, the proposed method effectively reduces blind spots in the interferogram while proserving detailed interferogram features, and the reconstructed spectrum is closer to the original signal. The effect of the algorithm is evaluated by using two quantitative analysis indexes, signal-to-noise ratio and mean square error. The results show that, compared with the principal component analysis algorithm and the average substitution algorithm, the signal-to-noise ratio of the reconstructed spectrum increases by 54.8% and 73.2%, respectively, while the MSE decreases by 32.2% and 99.8%, respectively.