Study On Visual Effect Optimization For Virtual Performance

The visual effect is one of the most important factors of the performance.Today’s performances are not limited to the real world,but also to areas such as computer graphics and virtual reality.The visual effect is influenced by many factors,such as camera control,stage lighting,etc.Generally speaking,the visual effect of performance is mainly designed by the designer according to the performance content,but in order to achieve a good performance effect,the designer needs to have enough experience.Therefore,if the camera effect and stage lighting effect can be adjusted in an automatic way,the designer’s energy can be greatly saved and the need for work experience can be reduced.At present,most of the methods of camera control optimization adopt the existing information in the scene or the action information of the role,and restrict the visual elements of the camera performance for a few targets through the overall analysis.However,this approach is not sufficient for real-time multiple targets stage performances.There are few researches on stage lighting so far,and the existing approach is to give users some choices to design by themselves.For stage performance,the lighting effect largely depends on the background music,so the background music can be analyzed to regulate the lighting effect.To solve the problems above,this paper starts from the perspective of automatic realization of stage performance visual effect optimization,researches and realizes the real-time stage camera control algorithm for multiple targets and music-driven stage lighting control algorithm.To be specific,the research work of this paper is as follows:(1)Propose a new camera space for multiple targets representation.It effectively solves the problem of less applicable targets in camera control.(2)Formulate the nonlinear optimization function,and adopt the model predictive control method to optimize the camera in real-time.It can make the algorithm more suitable for the real-time scene of stage performance.(3)Collect the 3D scene lighting color information and analyze it.Analyze the harmony degree of the scene lighting color creatively.(4)Classify the music,and adjust the lighting color based on the music tags and existing color information creatively.The experiments show that the method in this paper can effectively optimize the visual effect of the scene,and it has a certain reference significance for the visual performance optimization scheme of modern virtual performance.