PSNR (Peak Signal to Noise Ratio), which is the peak signal to noise ratio is the objective of evaluating the standard image, it has a limitation, a project is typically used between the maximum value signal and the background noise.
Wherein, the MSE X represents the current image and the reference image Y mean square error (MeanSquare Error), H, W are the height and width of the image; n is the number of bits per pixel, and generally 8, i.e., the number of pixels of the grayscale 256 . the PSNR unit is dB, larger value indicates less distortion.
PSNR is the most common and most widely used an image of objective evaluation, but it is based on an error between the corresponding pixel, i.e., based on image quality evaluation error sensitive. Since not take into account human visual characteristics (low spatial human eye contrast sensitivity higher frequency difference, the sensitivity of the human eye than the high luminance contrast difference in color, the human eye perceives the results of one region will be around near the affected area, etc.), the evaluation results and the subjective feeling of inconsistency so often.
40dB PSNR above described excellent image quality (i.e., very close to the original image),
the image quality is usually expressed 30-40dB good (i.e. distortion perceptible but acceptable),
the poor image quality described 20-30dB;
Finally, PSNR 20dB below the unacceptable image
description link: https: //blog.csdn.net/dxpqxb/article/details/85071338