Modified PSNR Metric for Watermarked-Image Assessment
Taha Basheer Taha · 2023
Digital images are exposed to different types of noise due to transformation, compression, or external data embedding. The assessment of these images is important to check their validity before being used in different applications. Because of its simplicity and straightforward calculations, the Peak Signal to Noise Ratio (PSNR) is used for evaluating digital images. PSNR, However, does not account for the human eye bias in preserving noise between textured and smooth areas. In this paper, a modified PSNR, named texture-based PSNR (T-PSNR), is proposed to enhance the PSNR metric for better image quality assessment. The metric is designed to be applied for watermarked image evaluation.