Study on Signal-to-noise ratio algorithms based on no-reference Image quality assessment

摘要

No-reference (NR) Image quality assessment is currently a hot research field of image quality evaluation. Since Signal-to-noise ratio (SNR) is an important indicator of image quality, some representative SNR evaluation algorithms based on NR are introduced in this paper, and are studied focusing on the consistency with subjective perception. The analysis results show that the algorithm based on perceptual masking of human visual system (HVS) has the best evaluation performance than other SNR evaluation algorithms, and the Frequency Domain method can be as the alternative one.