Design and performance of image quality models

Adequate image quality models should be applicable over a wide range of different types of pictures and distortions. Filtering and masking in conformity with known properties of the human visual system are important preprocessing components of such models. A first approach of an image quality model of this kind was given by Zetzsche and Hauske (1989). Further investigations by Xu (1997) showed that error segmentation and classification in connection with appropriate preprocessing lead to an image quality model with excellent performance. The role of the different stages of image quality models and existing limitations of their performance are be discussed