A new technique to inspect fabric defects based on biorthogonal wavelet lifting scheme is presented. In order to get fabric defect detection of rapidity and accuracy, the method is applied to constructs biorthogonal wavelet by lifting scheme and extracts the image characteristic features. The results show that its computing speed is increased by an average of 47% compared with the Mallat algorithm. And it overcomes the lack of space symmetry such as the orthogonal wavelet. The experimental results confirm that it can efficiently inspect four kinds of common fabric defects, warp-lacking, weft-lacking, oil stains, and holes with higher inspection speed.