[1]林俊杰 刘昊洋.基于GPU加速的多源点云融合算法研究[J].大众科技,2023,25(12):11-14.
 Research on Multi-Source Point Cloud Fusion Algorithm Based on GPU Acceleration[J].Popular Science & Technology,2023,25(12):11-14.
点击复制

基于GPU加速的多源点云融合算法研究()
分享到:

《大众科技》[ISSN:1008-1151/CN:45-1235/N]

卷:
25
期数:
2023年12
页码:
11-14
栏目:
信息技术与通信
出版日期:
2023-12-20

文章信息/Info

Title:
Research on Multi-Source Point Cloud Fusion Algorithm Based on GPU Acceleration
作者:
林俊杰 刘昊洋
(长安大学工程机械学院,陕西 西安 710064)
关键词:
GPU加速点云融合并行化计算
Keywords:
GPU acceleration point cloud fusion parallel computing
文献标志码:
A
摘要:
点云融合是一种将多个来源的点云数据配准和融合为一个更完整和准确的三维模型的重要算法。目前,已经出现了许多种点云融合算法,如串行编程方式(ICP)、非刚性配准等,这些算法通常需要大量的计算资源。文章基于图形处理单元(GPU)并行计算技术,提出了一种高效的多源点云融合算法,主要包括基于GPU的点云配准算法和点云融合算法。通过对点云数据在GPU中的存储和传输进行优化,以及设计高效的GPU并行算法,可以大幅度提高点云融合的计算速度和效率,为实际应用提供有力支持。
Abstract:
Point cloud fusion is an important algorithm in registering and fusing point cloud data from multiple sources into a more complete and accurate three-dimensional model. In the past few decades, many point cloud fusion algorithms have been proposed, such as ICP(In-circuit programmer) and non-rigid registration, etc. These algorithms typically require a large amount of computing resources. This article proposes an efficient multi-source point cloud fusion algorithm based on GPU(graphics processing unit) parallel computing technology, mainly including GPU based point cloud registration algorithm and point cloud fusion algorithm. By optimizing the storage and transmission of point cloud data in GPU and designing efficient GPU parallel algorithms, the computing speed and efficiency of point cloud fusion can be greatly improved, providing strong support for practical applications.

参考文献/References:

[1] 钟莹,张蒙. 基于改进ICP算法的点云自动配准技术[J]. 控制工程,2014,21(1): 37-40.[2] B?IS-SZOMOR?A, RIEMENSCHNEIDER H, VAN GOOL L, et al. Efficient volumetric fusion of airborne and street-side data for urban reconstruction[C]// 2016 23rd International Conference on Pattern Recognition (ICPR). Cancun, Mexico. IEEE, 2016: 3204-3209.[3] 刘鑫,许华荣,胡占义. 基于GPU和Kinect的快速物体重建[J]. 自动化学报,2012,38(8): 1288-1297.[4] 李健,曹垚,王宗敏,等. 融合k-means聚类和Hausdorff距离的散乱点云精简算法[J]. 武汉大学学报(信息科学版),2020,45(2): 250-257. [5] GUEVARA J, GENE-MOLA J, GREGORIO E, et al. 3D spectral graph wavelet point signatures in pre-processing stage for mobile laser scanning point cloud registration in unstructured orchard environments[J]. IEEE Sensors Journal, 2022, 22(2): 1720-1728.[6] 周春艳,李勇,邹峥嵘. 三维点云ICP算法改进研究[J]. 计算机技术与发展,2011,21(8): 75-77,81.[7] 冯文斌,刘宝华. 改进的SIFT算法图像匹配研究[J]. 计算机工程与应用,2018,54(3): 200-205,232.[8] 马瑞浩,朱枫,吴清潇,等. 基于图像分割的稠密立体匹配算法[J]. 光学学报,2019,39(3): 240-248.[9] LI S M, GE X M, HU H, et al. Laplacian fusion approach of multi-source point clouds for detail enhancement[J]. Journal of Photogrammetry and Remote Sensing, 2021, 171: 385-396.

备注/Memo

备注/Memo:
【收稿日期】2023-05-29【作者简介】林俊杰(1998-),男,福建莆田人,长安大学工程机械学院硕士研究生,研究方向为机器人自主导航。
更新日期/Last Update: 2024-03-04