Modeling Image Quality

Since digital images are subject to a wide variety of distortions during acquisition, processing, compression, storage, transmission and reproduction, it becomes necessary to have tools that make it possible to assess the Image Quality (IQ) during the whole production chain. By assessing the image quality during the image production stages make it possible to identify processing problems before the image is given/used by final users. For example, a printing office is required to print a given image (e.g. in a batch of brochures) as faithfully as requested by the customer. As another example, images that must be accessed on the Web often are compressed with a lossy algorithm (e.g. JPEG) to save bandwidth and download time. In order to retain the maximum quality and maximum compression ratio, image quality can be used to select the appropriate compression parameters. In both examples, image quality equates to faithfulness with an original one. We will see in the follows that other definitions of image quality exists and that they corresponds to other image properties. Image quality assessment can be done by manually subjective human rating or automatically by objective methods. This contribution aims to provide an overview of the state of the art of the Image Quality Assessment (IQA) methods reviewing the literature on objective image quality assessment, and classifying and summarizing the available metrics. BACKGROUND