An Image Quality Assessment Algorithm Based on Cross-correlation Coefficient with Edge Weight

In the practical application,image may need to be resized when to be observed,especially,in the circumstance of enlarging image resolution,the image quality would be degraded inevitably.In order to make image quality evaluation in this case,an image quality assessment algorithm based on prediction is proposed.The proposed algorithm use the cross-correlation coefficient between neighboring pixels as index.At the same time,considering the facts that a strong correlation between one pixel and its neighboring pixels would bring the fuzzy in the image edge area,so image is separated into different classification,different approach is adopted according to the different area.Meanwhile,in order to give attention of human visual system characteristics,the image multi-scale analysis is also to be embedded in the prediction algorithm.In the end,we get a more exact result score,and this score would be an index of distorted image quality score.The experiment results demonstrate that the proposed CCS algorithm can be used as a method to assess the image distorted quality,and increase the accuracy of image quality assessment comparing with conventional algorithms DIIVINE.