To address the shortcomings of the existing image quality assessment (I QA) methods,an IQA method based on deeply perception and learning for pooling ( DPLP) of visual features is proposed by deep learning theory modeling the charac teristics of human vision system (HVS).Firstly,according to spatial structure feature and multi-channel characteristics of the human eye,deeply visual infor m ation of image is successively processed by two-dimensional Log-Gabor wavelet tr ansform,gradient transform and spectrum residual,then each layer of vision inf ormation is extracted for quality assessment in order to increase the stability of image visual features.Secondly,the deeply pooling strategy is adopted for t he quality assessment information in order to overcome the uncertainty of HVS po oling.The first layer is the spatial pooling of the assessment map and the Gaus s weighting strategy which accords with the characteristics of human eye sensiti vity is adopted.The second layer is the frequency pooling of multi-channel ass e ssment and the learning-prediction strategy of BP neural network with HVS reaso n ing ability is adopted.The third layer is the assessment pooling of all levels of visual features and the regression function which holds adaptive characterist ics is adopted.Finally,experiments are carried out based on various types of d istortion images in reality.The results show that the proposed method has highe r consistency level between subjective assessment and objective assessment and b etter stability.