Predicting the quality of images compressed after distortion in two steps

In a typical communication pipeline, images undergo a series of processing steps that can cause visual distortion before being viewed. Given a high-quality reference image, a reference image quality assessment (IQA) algorithms can be applied after compression or transmission. However, the assumption of a high quality reference image is often not fulfilled in practice, contributing to less accurate quality predictions. Towards ameliorating this problem, we have devised a novel two-step image quality prediction approach that combines no-reference with reference quality measurements. Applying a first stage of no-reference IQA to determine the possibly degraded quality of the source image yields information that can be used to quality-modulate the reference prediction to improve its accuracy. We devise a simple and efficient weighted product of reference and no-reference stages that produces more reliable objective prediction scores. We also constructed a new dedicated database that is specialized for the design and testing of 'two-step' IQA models. Using this new resource, we show that two-step approaches yield outstanding performance when applied to compressed images whose original, pre-compression quality covers a wide range of realistic distortion types and severities. The two-step concept is versatile as it can use any desired reference and no-reference components. [...]Electrical and Computer Engineerin