Research On No-reference Authentic Image Quality Assessment Based On Subjective Perception

In the 21st century,human society is ushering into the information age of high speed development.Social media and rapid advances in camera and mobile device technology have led to the creation and consumption of a seemingly limitless supply of visual content.However,the vast majority of these digital images are captured by casual amateur photographers whose unsure hands and eyes often introduce annoying artifacts during acquisition.In addition,subsequent storage and transmission of visual media can further degrade their visual quality.Recent developments in visual modeling have elucidated the impact of visual distortions on perception of such pictures and videos.They have laid the foundation for automatic and accurate metrics that can identify and predict the quality of visual media as perceived by human observers.1 To address this problem,several objective blind or no-reference(NR)image quality assessment(IQA)algorithms have been developed to predict the perceptual quality of a given(possibly distorted)image without additional information.Such quality metrics could be used to monitor and control multimedia services on networks and devices or to prioritize quality of transmission over speed.Real-world images are usually afflicted by mixtures of distortions that differ significantly from the single.For the real distortion,image quality assessment will be more difficult.So we focus on the problem,research single distortion image quality evaluation method firstly,on the basis of above research real distortion image quality evaluation method.The research achievements of this paper are as follows:1)In this paper,we propose a simple and efficient image quality assessment algorithm for authentic distortion without reference image.The algorithm extract respectively the image's information entropy in the spatial domain and spectral domain,LBP characteristics of gradient domain,improved HOG features as the image's distortion degree of characterization based on the natural scenes statistical of image.After feature extraction,using the PC A method to reduce the dimension of data and then do the standardization of data processing.XGBoost is used for model regression.The experimental results show that the tree assemble model has good consistency with human subjective perception.They also show that the algorithm has good generality.Compared to the previous classic methods,the algorithm is faster.2)In this paper,a kind of deep convolution neural network architecture is proposed as a method to capture the depth local and global information of the image.Systematically assess the performance of the convolutional neural network to predict objective image quality scores.Experiments show that the method is effective to simulate the human perception and cognitive mechanism.This paper proposes a deep learning model to classify the quality level of the image,and introduce a method to extend the database ten times.What' more,improve the accuracy of the classification of image quality level.