A novel free reference image quality metric using neural network approach
A.chetouani, A.Beghdadi, S.Chen and G.Mostafaoui, Int. Workshop on Video Process. Qual. Metrics Cons., Electronics, pp:1-4, Jan. 2010.
In this paper, three Non reference(NR) metric is fused in a Neural Network to measure the image quality. The focus is on blocking, blurring and ringing artifacts which is caused by compression. In order to measure the quality of the image, the features are extracted and given as input to ANN. It uses a MLP with one hidden layer and one output. It uses sigmoid activation function to describe the relation between the subjective and subjective scores. The scores are scaled in the range [-1,+1] where -1 indicates the best quality.