Application and Performance Analysis of the Improved Quadric Error Metrics Algorithm in 3D Mesh Simplification
Wenxiang Yuan, Qi Liu, Lulu Zhao, Yu Deng, Jing Hu, Jiarui Hai, Shengjie Wang, Yuanchi Huang, Wenyi Ge · 2025
With the rapid development of information technology, the simplification of 3D mesh models has become an important means to enhance rendering efficiency and real-time interactive performance. This paper proposes an improved method based on the Quadric Error Metrics (QEM) algorithm, aiming to reduce errors during the simplification process, prevent the degradation of feature regions, and ensure the coherence and consistency of visual effects. The improved algorithm introduces local mesh feature measurements and combines geometric feature analysis to better preserve model details during simplification. Experimental results show that, when applied to 3D meshes of deicing vehicles and aircraft in airport scenarios, this method demonstrates significant advantages over traditional QEM and vertex clustering algorithms in terms of Hausdorff distance and mean error, effectively reducing geometric deviations and improving the quality of simplified models. In addition, this paper employs bilateral filtering for mesh saliency computation, further optimizing simplification effects. The algorithm is suitable for applications requiring high fidelity and holds significant practical value. Future research may further explore the integration of human visual perception models with simplification algorithms and promote the development of multi-resolution mesh structures to meet the dual demands of performance and accuracy in real-time rendering.