Optimisation of 3D models for visualisation in virtual reality systems

The relevance of the study is conditioned by the growing use of virtual reality in education, medicine, design and engineering systems, where traditional 3D models are often too resource-intensive for stable interactive reproduction. The purpose of the study was to identify and test the effectiveness of complex adaptation of digital objects to interactive spatial reproduction in conditions of limited hardware resources. The research methodology combined theoretical analysis with experimental testing in Unity 2022.3 on the Meta Quest 2 platform with performance and visual quality assessment. The results of the study showed that direct use of traditional 3D models in virtual reality is ineffective due to excessive load on hardware resources and instability of scene reproduction. It was established that the highest efficiency is provided by complex adaptation, which combines simplification of the polygonal grid, reduction of the volume of texture data, simplification of materials, application of detail levels, and stage optimisation. After adaptation, the number of polygons in the studied models decreased by 95-96%, and the volume of texture data – by 85-88%. In the test virtual environment, the average frame rate increased from 42 to 85 frames per second, which confirmed a significant increase in scene performance. However, the increase in performance was not accompanied by a critical loss of visual quality, since the overall shape of objects remained at the level of 98-99%, and the level of visual immersion remained high. The results of the study can be used by developers of virtual reality applications, computer graphics specialists, teachers, designers, and engineers to create training simulators, architectural visualisation, medical simulators, and technical virtual reality systems with increased performance and visual realism