Identifying the image quality measures that have highest sensitivity to distortions would help systematic design of coding, communication and imaging systems and of improving or optimizing the picture quality for a desired quality of service at a minimum cost. Image quality measurement basically consists of two approaches: Subjective measurements and Objective measurements.
Figure 1 shows the classification of IQA measures. Subjective measurements are the result of human experts providing their opinion of the image quality and objective measurements are performed with mathematical algorithms. The goal of research in objective image quality assessment is to develop quantitative measures that can automatically predict perceived image quality (VQEG, 2000). An objective image quality metric can play a variety of roles in image processing applications. It can be used to dynamically monitor and adjust image quality. It can be used to optimize algorithms and parameter settings of image processing systems. It can be used to benchmark image processing systems and algorithms (Pappas et al.,2000).
Classification of image assessment
In full-reference image quality assessment methods, the quality of a reproduced image is evaluated by comparing it with the original image that is assumed to have perfect quality. In reduced-reference image quality assessment methods, limited information are available for both reproduced image and original image to evaluate the quality of the test images. No-reference metrics try to assess the quality of an image without any reference to the original one.