目的 为了解决数字印刷模糊混合失真图像质量评价算法与人眼视觉主观评价一致性较低的问题。方法 基于SSEQ算法思想在空间域和频率域中提取混合失真图像的信息熵、亮度方差、频域熵以及高频系数奇异值改变量关键特征，利用支持向量机训练构建印刷模糊混合失真图像的无参考评价模型。结果 客观验证了25幅图像，算法客观计算得分与主观评价得分的SROCC系数为0.94，PLCC系数为0.93，主客观一致性较高。结论 所提算法与人眼视觉感知具有很好的主观一致性，可以用于图像等印刷品的质量客观评价。

The work aims to solve the problem that the blurred image of digital printing multiply distorted quality assessment algorithm has low consistency with the subjective evaluation of human perception. Key characteristics including entropy, brightness variance, frequency entropy and the high frequency coefficient singular value change were extracted from distorted image in spatial and frequency domain based on the idea of SSEQ algorithm to establish the no-reference image quality assessment method with support vector machine in this paper. 25 images were verified objectively and the SROCC coefficient and PLCC coefficient of the objective calculation score and the subjective evaluation score of the algorithm were 0.94 and 0.93 respectively, with high subjective and objective consistency. The algorithm is in good subjective consistency with the ocular perception and suited for the printed image quality assessment.