A spectral reflectance reconstruction algorithm was studied in the paper, in order to improve the accuracy of the reconstructed spectral reflectance of the colors of different objects. The system response values of each color in the test sample were obtained by high-precision multi-spectral imaging system. And the spectral reflectance values of the colors were measured by spectrophotometer. The spectral reconstruction of the sample data was conducted with Wiener, Adaptive Wiener and the proposed algorithms, and the reconstruction results of the three algorithms were evaluated. The experiment results showed that the root-mean-square-error of the proposed method was 0.0355, and the average color difference was 1.4349, outperforming the other two algorithms in terms of both spectral and colorimetric accuracy. In conclusion, the proposed method had effective performance in improving the spectral reconstruction precision, and can be applied to multi-spectral imaging reproduction.