1. Software version
MATLAB2013b
2. Theoretical knowledge of this algorithm
A non-reference evaluation algorithm of image quality is carried out through fuzzy neural network. Through the training of fuzzy neural network, the functional relationship between image feature value and image differential mean subjective evaluation value (DMOS) is established. GRNNs are often used for function approximation. It has a radial basis hidden layer and a special linear layer. The number of neurons in both the first and second layers is equal to the number of input sample vector pairs. The structure of GRNN is shown in Figure 2-3. The entire network includes four layers of neurons: input layer, pattern layer, summation layer and output layer.
For an unknown probability density function f (x, y), it can be estimated nonparametrically from the observed samples of x and y: