Visual Effect Optimization Algorithm of Virtual Performance Based on PDE Model

With the rapid development of computer graphics and artificial intelligence technology, virtual performance has become an important part of the digital entertainment industry. In order to solve the inefficiency and inaccuracy of manual design and experience adaptation in optimizing the visual effects of virtual performances, this paper optimizes the visual effects based on partial differential equations (PDE) and realizes automatic adjustment of the visual effects of virtual performances. First, the background of visual optimization of virtual performances is analyzed. Next, this paper designs an optimization algorithm based on the PDE model to simulate the dynamic image development process and realize real-time adjustment of camera recording effects and scene lighting. Finally, experimental verification shows that the optimization algorithm based on the PDE model proposed in this paper significantly improves the visual effects of virtual performances. The experimental results show that the Delta E values of the proposed algorithm are higher than those of the PDE model when compared with the traditional manual adjustment method in practice. The former has a smaller fluctuation, maintained at 7.3–9, while the latter has a larger fluctuation, maintained at 2.5–3, which indicates that the PDE model is more efficient and stable in color accuracy.