An image quality assessment algorithm based on feature selection

摘要: Image Quality Assessment(IQA) is of fundamental importance to numerous imaging and video processing applications. For most the applications, perceptual meaningful measure one which can automatically assess quality images or videos in a perceptually consistent manner. However, commonly used IQA metrics are not well with human judgments image quality. Recently, SSIM metric takes people's visual characteristics into consideration performs much better than traditional PSNR/MSE. But defects it still exit on some specific kinds distortions. A new algorithm based feature selection proposed this paper. Local gradient entropy phase congruency added framework. Through in-depth definition plus pooling strategy, LIVE datasets.