| 基于噪声点云的三维场景重建方法 |
| |
| 引用本文： | 张健,李新乐,宋莹,王仁,朱凡,赵晓燕.基于噪声点云的三维场景重建方法[J].计算机工程与设计,2020,41(4):972-977. |
| |
| 作者姓名： | 张健 李新乐 宋莹 王仁 朱凡 赵晓燕 |
| |
| 作者单位： | 中国航天科工集团第二研究院 706 所 ,北京 100854;华中科技大学计算机科学与技术学院,湖北武汉430074 |
| |
| 基金项目： | 装备预研项目;十三五全军共用信息系统装备专用技术预先研究基金项目;创新基金 |
| |
| 摘 要： | 为解决深度相机数据噪声导致的过平滑、细节部分精度不够等问题,提出一种基于点云的三维重建方法。采用基于带概率模型的点云设计全局的位姿估计方法用于模型的表达与聚合;设计基于场景特征的子图划分方法,减少冗余子图,使用基于点云的方法进行位姿图优化,实现子图构建;采用基于GPU加速的FPFH算法,从点云上提取特征点进行匹配,提高闭环检测实时性。实验结果表明,该方法可以显著提高模型的质量,在闭环检测上有更高的鲁棒性。
|
| 关 键 词： | 深度相机 三维建模 点云 子图 闭环检测 |
3D reconstruction method based on point cloud with noise |
| |
| ZHANG Jian,LI Xin-le,SONG Ying,WANG Ren,ZHU Fan,ZHAO Xiao-yan.3D reconstruction method based on point cloud with noise[J].Computer Engineering and Design,2020,41(4):972-977. |
| |
| Authors: | ZHANG Jian LI Xin-le SONG Ying WANG Ren ZHU Fan ZHAO Xiao-yan |
| |
| Affiliation: | (Institute 706,Second Academy of China Aerospace Science and Industry Corporation,Beijing 100854,China;Department of Computer Science and Technology,Huazhong University of Science and Technology,Wuhan 430074,China) |
| |
| Abstract: | To solve the reconstruction problems,such as over-smoothed geometric details and the low precise detail,caused by the noisy data obtained using the depth camera,a 3D reconstruction method based on point cloud was proposed.A global pose estimation method based on the point cloud with probability model was proposed for the representation and fusion of the model.A method of inserting subgraphs was proposed to reduce redundant subgraphs.The fast point feature histogram(FPFH)based on the graphics processing unit(GPU)was used to improve the real-time performance of loop detection.Experimental results show that the proposed model can significantly improve model’s quality and has higher robustness in loop detection. |
| |
| Keywords: | depth camera 3D reconstruction point cloud subgraph loop detection |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
