Optimization of building model based on 5G virtual reality technology in computer vision software

The 5G virtual reality system needs to interact with the user to draw the scene in real time. The contradiction between the complexity of the scene model and the real-time interaction is the main problem in the operation of the virtual reality system. The model optimization strategy of architectural scene in virtual reality design is studied, and the method of architectural scene model optimization is summarized. This article aims to study the optimization of computer vision software modeling through 5G virtual reality technology. In this paper, the optimization of the architectural model is studied by the method of image gray scale transformation, computer vision detection technology and virtual modeling technology. The four experiments are comprehensive evaluation and quantitative evaluation, comparison of channel estimation performance of different pilot structures, comparison of calculated and true values of external azimuth elements, and the effect of window-to-wall ratio on energy consumption per unit of residential building. The results show that hollow bricks of building materials have a great impact on the environment. The values of the three pixel coordinates X, Y, and Z calculated by the unit quaternion method are 1.27, 1.3, and -6.11, respectively, while the actual coordinate positions are 1.25, 1.37, and -6.22, respectively. It can be seen that the outer orientation element value calculated by the quaternion-based spatial rear intersection method is not much different from the actual value, and the correct result can be accurately calculated.