Three-Dimensional Reconstruction and Virtual Display Technology of Cultural Heritage Based on Computer Vision

In the field of cultural heritage protection and display, due to limitations in technological means, traditional research methods (manual restoration) are difficult to fully display the details and historical context of cultural relics, and may also damage cultural relics in physical protection. This article combines modern Three-Dimensional (3D) scanning with virtual reality to explore a new way of cultural heritage reconstruction and protection. This article uses advanced laser scanners, allowing operators to scan the measured cultural relics from different angles, thereby generating high-precision point clouds. Each point accurately records the coordinate values of a specific object surface. The collected point cloud data was denoised by data cleaning (excluding non target point clouds), point cloud stitching (integrating multi-view point clouds to form integrated 3D data), and optimization (improving data accuracy and availability). Based on the point cloud data, advanced 3D reconstruction algorithms such as multi-view geometry and structured light were used to establish an overall 3D model. The physical data of item 1 is 20.5 m, the 3D model data is 20.48 m, and the evaluation score is 9 points; The physical data of item 2 is 50.0 m, the 3D model data is 50.05 m, and the evaluation score is 8 points. This article is based on the three-dimensional reconstruction and virtual display technology of cultural heritage using computer vision, which is beneficial for improving the authenticity and public participation of cultural heritage.