Research on LIDAR-based 3D Mapping Method for Indoor Smoke Scene
When an indoor fire occurs,LiDAR sensors can be used to construct a map of accident scene to assist in fire rescue.Due to the enclosed nature of the interior,a large number of smoke particles accumulate in the environment,causing noise in LiDAR sensor and resulting in distorted maps.To solve this problem,traditional point cloud filtering algorithm is proposed by using a radius filter assisted by dynamic threshold and intensity information to remove smoke noise from LiDAR point cloud.After that,feature extraction and nearest neighbor matching are performed on point cloud for pose estimation and map construction.An indoor smoke scene for data acquisition and algorithm testing experiments is built,the results show that the proposed method can effectively eliminate the influence of smoke on LiDAR 3D mapping and has a good retention of non-noise points.
LiDAR mappingindoor mappingpoint cloud filteringsmoke scene
万方数据
维普
