Algorithm for Image Quality Assessment Based on Color Features

As the most direct and efficient features to describe images,color features currently become one of the most significant parameters in image quality assessment.Based on the theory of structural similarity,a model of full-reference objective quality assessment system for image on the basis of color features is proposed,which takes the affect of chrominance on the human visual perception into consideration.After the theoretical analysis,a new similarity model using color moments of partitions is established in this paper.Experimental results show that the model has high consistency with the subjective results.