Non-reference image quality evaluation method based on HVS characteristics

In order to effectively measure the image quality of different distortion types, a non-reference image quality evaluation algorithm based on natural scene statistics (NSS) model is proposed. This algorithm uses the statistical characteristics of the normalized brightness coefficients of natural images to tend to obey the characteristics of the generalized Gaussian probability distribution. Firstly, the regions of interest and non-interest regions of natural images are extracted, and then the image features are extracted in the corresponding regions, and finally realized image quality evaluation method based on visual regions of interest. Based on the LIVE2 image database, a large number of experiments are performed and compared with other classic algorithms. Experimental results show that the improved algorithm in this paper has high subjective consistency in the evaluation results of degraded images with multiple types of distortion coexisting, and can accurately reflect the visual perception quality of the human eye.