This paper proposes an information content weighting approach for perceptual image quality assessment (IQA). The authors suggest that optimal perceptual weights for pooling in IQA algorithms should be proportional to local information content, which can be estimated using advanced statistical models of natural images. They tested this hypothesis using six publicly available image databases and found three key findings: (1) information content weighting consistently improves IQA performance; (2) even the widely criticized PSNR can be converted into a competitive perceptual quality measure with information content weighting; and (3) the best overall performance is achieved by combining information content weighting with multiscale structural similarity measures (SSIM). The paper discusses various pooling strategies, including Minkowski pooling, local quality/distortion-based pooling, saliency-based pooling, and object-based pooling. The authors then introduce an information content weighting method based on the Gaussian scale mixture (GSM) model, which quantifies local information content in units of bit. This method is combined with multiscale SSIM to create an information content weighted SSIM (IW-SSIM) algorithm. The authors also propose an information content weighted PSNR (IW-PSNR) algorithm. The proposed algorithms are validated against 13 other IQA algorithms on six publicly available image databases. The results show that IW-SSIM achieves the best overall performance, outperforming other methods in terms of correlation coefficients and error metrics. The authors conclude that the optimal pooling weights should be proportional to local information content, as measured in units of bit. The paper also discusses the potential for future research, including the use of more advanced image models and the incorporation of color components in IQA.This paper proposes an information content weighting approach for perceptual image quality assessment (IQA). The authors suggest that optimal perceptual weights for pooling in IQA algorithms should be proportional to local information content, which can be estimated using advanced statistical models of natural images. They tested this hypothesis using six publicly available image databases and found three key findings: (1) information content weighting consistently improves IQA performance; (2) even the widely criticized PSNR can be converted into a competitive perceptual quality measure with information content weighting; and (3) the best overall performance is achieved by combining information content weighting with multiscale structural similarity measures (SSIM). The paper discusses various pooling strategies, including Minkowski pooling, local quality/distortion-based pooling, saliency-based pooling, and object-based pooling. The authors then introduce an information content weighting method based on the Gaussian scale mixture (GSM) model, which quantifies local information content in units of bit. This method is combined with multiscale SSIM to create an information content weighted SSIM (IW-SSIM) algorithm. The authors also propose an information content weighted PSNR (IW-PSNR) algorithm. The proposed algorithms are validated against 13 other IQA algorithms on six publicly available image databases. The results show that IW-SSIM achieves the best overall performance, outperforming other methods in terms of correlation coefficients and error metrics. The authors conclude that the optimal pooling weights should be proportional to local information content, as measured in units of bit. The paper also discusses the potential for future research, including the use of more advanced image models and the incorporation of color components in IQA.