Abstract: Wavelet-based contourlet transform (WBCT) can efficiently reflect visual characteristics of iris textures. After localizing a region of evaluation by using a cascaded BP neural network, the region of image is decomposed with WBCT. Then, 5 image quality assessment indices are defined and calculated for defocused images, motion blurred images, contact lens wear images, eyelash and eyelid occluded images, respectively. Experimental results indicate that these indices can precisely evaluate these cases with low computational cost. Moreover, the evaluation results have good consistency with subjective assessment of human beings. Compared with the existing methods, our method achieves better performance.