An intelligent perception and data processing model for building information based on LiDAR and UNet

The maturity of intelligent algorithms in the fields of computer vision and 3D data processing has provided technical support for building information processing. However, traditional processing methods for building information are inefficient and tend to overlook the semantic context correlation and humanistic attributes of building information. Therefore, a research proposal has been made to develop an intelligent perception and data processing model for building information that integrates LiDAR, U-shaped networks, Transformer, and graph neural networks. Experimental results show that the structural integrity reasoning accuracy of the proposed model reaches 91.38% and the F1 score is 91.64%. The above experimental results indicate that the proposed model can efficiently extract building semantic information and spatial correlation information, providing scientific solutions for 3D building reconstruction, precise restoration, and intelligent building operation and maintenance.