No-reference Image Blur Assessment Method Based on Local Perceive and Visual Saliency

This paper proposes a new image quality assessment method based on visual saliency and local perceptual sharpness. Firstly, the effective visual saliency image salient region is detected, and then the image local variance and mean construct effective image saliency weights are combined. Secondly, the image sharpness of local spectrum feature of local spatial features and sharpness are respectively computed, and then a new feature sharpness map is formed by combing the visual saliency weight of the image. Finally, the final evaluation index value of image quality is computed by the analysis of the sharpness map.Some experimental results show that the proposed method is effective than traditional evaluation method based on local perceptual sharpness in the classical database TID2008, CSIQ and LIVE because we effectively combine with color information and image saliency and improve the quality evaluation.