Measuring the Blurring Artifact in Grayscale Images Compressed with a Wavelet Transform

摘要

Research into image quality evaluation aims to define a single value to describe the difference from the original image. In this report, we compare quality measures obtained from various sources. They are all based on the degradation of structural information in the image. Detection of the changes in the structural information is considered as a typical feature of the human visual system. Another requirement for a quality measure comes from the computational complexity, which should be kept relatively low. Our experiments indicate that the existing measures should be developed further to better capture the blurring artifact.