A fragment detection method for static explosion field based on 3D reconstruction

DOI:	10.11809/bqzbgcxb2024.06.028
Keywords:	three dimensional reconstruction; point cloud segmentation; adaptive threshold; region growing; normal vector
Abstract:	In order to solve the problem of low efficiency of statistical analysis of fragments landing in static explosion test in shooting range, this paper proposes a fragment detection method based on adaptive threshold color region growth of multi view Iterative reconstruction. By introducing adaptive thresholds in point cloud segmentation, the detailed information of 3D point clouds is effectively preserved, enhancing the algorithm’s generalization ability for different scenes in the shooting range. In response to the problem of poor efficiency in dealing with a large number of point clouds, the point cloud normal vector is used for color region color growth segmentation, effectively reducing detection errors and missed detections, and improving the accuracy of fragment detection. The experimental results show that the proposed detection method has a detection accuracy of 95%, and the error of fragment impact point is less than 4%. It can effectively segment the target parameters of fragments, providing experimental basis for subsequent fragment statistical analysis work.
Published:	2024-06-28
PDF