A Novel Game Graphics Quality Evaluation Model Using Saliency and Resolution Information
Binlin Feng, Ying Chu, L. P. Zhou, Hengyong Yu · 2025
With the advancement of computer graphics techniques, the gaming industry has generated a growing array of computer graphics images (CGIs) featuring complex and various textures that are shown on devices with different resolution capacities. However, some CGIs encounter graphic distortions due to limitations in rendering precision that substantially diminishes user experiences. It is important to evaluate the quality of CGIs in an effective way. Meanwhile, current assessment techniques are not good enough because game pictures have intricate textures and considerable resolution variations. In this paper, a novel blind image quality assessment (IQA) method is proposed, which employs image saliency and resolution information to improve its ability to detect the quality of distorted texture features. Experimental results in game-related IQA datasets demonstrate enhanced prediction accuracy, robustness and generalizability of the proposed method.