Structural Similarity Based Image Quality Assessment

It is widely believed that the statistical properties of the natural visual environ-ment play a fundamental role in the evolution, development and adaptation of the human visual system (HVS). An important observation about natural image sig-nals is that they are highly structured. By “structured signal”, we mean that the signal samples exhibit strong dependencies amongst themselves, especially when they are spatially proximate. These dependencies carry important information about the structure of the objects in the visual scene. The principle hypothesis of structural similarity based image quality assessment is that the HVS is highly adapted to extract structural information from the visual field, and therefore a measurement of structural similarity (or distortion) should provide a good approx-imation to perceived image quality. In this chapter, we introduce structural similarity as an alternative design phi-losophy for objective image quality assessment methods. This is different from and complementary to the typical HVS-based approaches, which usually calculate signal difference between the distorted and the reference images, and attempt to quantify the difference “perceptually ” by incorporating known HVS properties. 1.1 Structural Similarity and Image Quality In full-reference image quality assessment methods, the quality of a test image is evaluated by comparing it with a reference image that is assumed to have perfec