基于多视角边缘检测融合的放射性物品运输容器特征提取

张沛东; 肖连博; 张煜航; 焦力敏; 孙洪超; 庄大杰; 李国强

doi:10.19554/j.cnki.1001-3563.2025.11.032

PDF(10400 KB)

包装工程（技术栏目） ›› 2025, Vol. 46 ›› Issue (11) : 294-302. DOI: 10.19554/j.cnki.1001-3563.2025.11.032

装备防护

基于多视角边缘检测融合的放射性物品运输容器特征提取

张沛东, 肖连博, 张煜航, 焦力敏, 孙洪超, 庄大杰, 李国强

作者信息 +

Feature Extraction of Radioactive Transport Container Based on Multi-view Edge Detection Fusion

ZHANG Peidong, XIAO Lianbo, ZHANG Yuhang, JIAO Limin, SUN Hongchao, ZHUANG Dajie, LI Guoqiang

Author information +

文章历史 +

摘要

目的针对放射性物品运输容器可靠性评估中存在的三维形变测量精度不足的瓶颈问题,提出一种融合多视角深度图投影与亚像素边缘检测的三维点云特征提取方法,以高精度几何特征数据支撑容器的结构优化设计与安全验证。方法首先,构建最小包围立方体,并设计六自由度多视角正交投影系统,将三维点云降维映射至二维深度图,实现容器全表面覆盖;其次,基于拟合边缘模型及局部灰度面积效应,检测亚像素边缘点;进一步结合投影参数化映射与动态曲率插值算法,引入法向量连续性约束及加权融合策略,重构高保真三维特征点;最后,搭建实验系统,获取容器三维点云数据,基于简化率、配准误差及计算时间指标,并与现有点云特征提取方法进行对比。结果采用所提方法获取特征点云,其配准最大平均误差为1.33 mm,最大均方根误差为1.39 mm,简化率为1.6%。与其他文献的结果相比,所提方法的简化率更小,计算时间更短,且误差更小。结论所提方法有效解决了大规模点云处理与几何特征保真之间的矛盾,显著提升了放射性容器三维特征提取的精度和效率,为核能工程安全评估提供了可靠的几何形变量化分析技术支撑。

Abstract

The work aims to propose a 3D point cloud feature extraction method combining multi-view depth map projection and sub-pixel edge detection, and support the structural optimization design and safety verification of containers with high-precision geometric feature data, so as to solve the bottleneck problem of insufficient accuracy of 3D deformation measurement in reliability evaluation of radioactive materials transport containers. Firstly, by constructing the minimum bounding cube and designing a six-degree-of-freedom multi-view orthogonal projection system, the three-dimensional point cloud was mapped to a two-dimensional depth map to realize the full surface coverage of the container. Secondly, based on the fitting edge model and local gray area effect, sub-pixel edge points were detected; Furthermore, combining projection parametric mapping with dynamic curvature interpolation algorithm, normal vector continuity constraint and weighted fusion strategy were introduced to reconstruct high-fidelity three-dimensional feature points; Finally, an experimental system was built to obtain the 3D point cloud data of the container, and compared with the existing point cloud feature extraction methods based on simplification rate, registration error and calculation time. The experimental results showed that the maximum average error, the maximum root mean square error and the simplification rate of the feature point cloud obtained by the proposed method were 1.33 mm, 1.39 mm and 1.6%, respectively. Compared with other literature results, the proposed method had a smaller simplification rate, faster calculation time and smaller error. The method proposed in this study effectively solves the contradiction between large-scale point cloud processing and geometric feature fidelity, significantly improves the accuracy and efficiency of three-dimensional feature extraction of radioactive containers, and provides reliable technical support for geometric deformation quantitative analysis for nuclear energy engineering safety assessment.

导出引用

张沛东, 肖连博, 张煜航, 焦力敏, 孙洪超, 庄大杰, 李国强. 基于多视角边缘检测融合的放射性物品运输容器特征提取[J]. 包装工程（技术栏目）. 2025, 46(11): 294-302 https://doi.org/10.19554/j.cnki.1001-3563.2025.11.032

ZHANG Peidong, XIAO Lianbo, ZHANG Yuhang, JIAO Limin, SUN Hongchao, ZHUANG Dajie, LI Guoqiang. Feature Extraction of Radioactive Transport Container Based on Multi-view Edge Detection Fusion[J]. Packaging Engineering. 2025, 46(11): 294-302 https://doi.org/10.19554/j.cnki.1001-3563.2025.11.032

中图分类号： TB487

参考文献

[1] 孙谦, 庄大杰, 孙洪超, 等. 核燃料组件运输容器应用现状概述[J]. 包装工程, 2022, 43(13): 142-150.
SUN Q, ZHUANG D J, SUN H C, et al.Overview of Application Status for Nuclear Fuel Assembly Transport Package[J]. Packaging Engineering, 2022, 43(13): 142-150.
[2] KIM K S, KIM J S, CHOI K S, et al.Dynamic Impact Characteristics of KN-18 SNF Transport Cask - Part 1: An Advanced Numerical Simulation and Validation Technique[J]. Annals of Nuclear Energy, 2010, 37(4): 546-559.
[3] BARTOL K, BOJANIĆ D, PETKOVIĆ T, et al.A Review of Body Measurement Using 3D Scanning[J]. IEEE Access, 2021, 9: 67281-67301.
[4] YAO Z W, ZHAO Q X, LI X F, et al.Point Cloud Registration Algorithm Based on Curvature Feature Similarity[J]. Measurement, 2021, 177: 109274.
[5] WANG G L, WU L S, HU Y, et al.Point Cloud Simplification Algorithm Based on the Feature of Adaptive Curvature Entropy[J]. Measurement Science and Technology, 2021, 32(6): 065004.
[6] XU Y S, TONG X H, STILLA U.Voxel-Based Representation of 3D Point Clouds: Methods, Applications, and Its Potential Use in the Construction Industry[J]. Automation in Construction, 2021, 126: 103675.
[7] XIE Z P, LANG Y C, CHEN L Q.Geometric Modeling of Rosa Roxburghii Fruit Based on Three-Dimensional Point Cloud Reconstruction[J]. Journal of Food Quality, 2021, 2021(1): 9990499.
[8] SUN Y, ZHANG S H, WANG T Q, et al.An Improved Spatial Point Cloud Simplification Algorithm[J]. Neural Computing and Applications, 2022, 34(15): 12345-12359.
[9] LEAL E, SANCHEZ-TORRES G, BRANCH-BEDOYA J W, et al. A Saliency-Based Sparse Representation Method for Point Cloud Simplification[J]. Sensors, 2021, 21(13): 4279.
[10] GUMHOLD S, WANG X, MACLEOD R S.Feature Extraction from Point Clouds[C]//IMR, 2001: 293-305.
[11] HO H T, GIBBINS D.Curvature-Based Approach for Multi-Scale Feature Extraction from 3D Meshes and Unstructured Point Clouds[J]. IET Computer Vision, 2009, 3(4): 201-212.
[12] WANG L, YUAN B.Feature Point Detection for 3D Scattered Point Cloud Model[J]. Signal Process, 2011, 27(6): 932-938.
[13] WANG S Q, HU Q J, XIAO D S, et al.A New Point Cloud Simplification Method with Feature and Integrity Preservation by Partition Strategy[J]. Measurement, 2022, 197: 111173.
[14] POTAMIAS R A, BOURITSAS G, ZAFEIRIOU S.Revisiting Point Cloud Simplification: A Learnable Feature Preserving Approach[M]//Computer Vision - ECCV 2022. Cham: Springer Nature Switzerland, 2022: 586-603.
[15] WU J S, LAI X M, CHAI X L, et al.Feature-Based Point Cloud Simplification Method: An Effective Solution for Balancing Accuracy and Efficiency[J]. The Journal of Supercomputing, 2024, 80(10): 14120-14142.
[16] ALSHAWABKEH Y.Linear Feature Extraction from Point Cloud Using Color Information[J]. Heritage Science, 2020, 8(1): 28.
[17] HU H L, LI Z, JIN X G, et al.Curve Skeleton Extraction from 3D Point Clouds through Hybrid Feature Point Shifting and Clustering[J]. Computer Graphics Forum, 2020, 39(6): 111-132.
[18] SUN Y, ZHANG S H, WANG T Q, et al.An Improved Spatial Point Cloud Simplification Algorithm[J]. Neural Computing and Applications, 2022, 34(15): 12345-12359.
[19] DING Z H, ZHENG S H, ZHANG F L, et al.Lensfree Auto-Focusing Imaging with Coarse-to-Fine Tuning Method[J]. Optics and Lasers in Engineering, 2024, 181: 108366.
[20] TRUJILLO-PINO A, KRISSIAN K, ALEMÁN-FLORES M, et al. Accurate Subpixel Edge Location Based on Partial Area Effect[J]. Image and Vision Computing, 2013, 31(1): 72-90.
[21] DU H, HU Z Y, LU D P, et al.Density Clustering Algorithm Based on KD-Tree and Voting Rules[J]. Computers, Materials & Continua, 2024, 79(2): 3239-3259.
[22] QI J K, HU W, GUO Z M.Feature Preserving and Uniformity-Controllable Point Cloud Simplification on Graph[C]//2019 IEEE International Conference on Multimedia and Expo (ICME). IEEE, 2019: 284-289.