Algorithm for fire detection and 3D point cloud data fusion construction under deep learning

Fire in high-rise buildings is a major hidden danger that threatens people's lives, and in order to know the fire situation, this paper proposes to utilize the optimized YOLOv11 model to accurately identify the flames and smoke, and to use 3D LIDAR to construct point cloud maps of high-rise buildings in real time. Through the high-definition camera and 3D LiDAR to achieve the fire and environment map construction, and the fire information mapped to the 3D map to achieve the three-dimensional visualization of fire information. The experimental results show that the improved algorithm improves the accuracy and F1 score by 1 percentage point compared with the original, and the fused point cloud map has high imaging fineness and complete information.