Objective image quality measure based on Weber-weighted mean absolute error

This paper introduces three simple image quality measures based on Weber law. After normalizing the original and distorted images, the absolute error between them is found. The final index is a pixel by pixel weighted sum of errors. The maximum between the original image pixel and its negative is the inverse of the weighting factor. Results indicate a better correlation with human perception compared to conventional mean square error MSE. Despite the simplicity of the proposed model, it compares well with other complex schemes in the literature and its complementary outcome is promising in improving those schemes.