基于高斯分布和多元正态分布结合模型的点云异常点检测与去噪方法研究  

Enhanced point cloud abnormality detection and denoising method utilizing combined Gaussian and multivariate normal distribution model

作　　者：李舒雯 周思瑶 曲畅 LI Shuwen;ZHOU Siyao;QU Chang(College of Computer Science and Cyber Security,Chengdu University of Technology,Chengdu 610059,China;College of Mathematics and Physics,Chengdu University of Technology,Chengdu 610059,China)

机构地区：[1]成都理工大学计算机与网络安全学院,成都610059 [2]成都理工大学数理学院,成都610059

出　　处：《计算机应用文摘》2024年第20期194-196,共3页

摘　　要：针对异常点检测在点云采集中,由于复杂数据分布导致去噪准确率和效率下降的问题,文章提出了一种结合高斯分布模型和多元正态分布模型的异常点检测方法。实验结果表明,该方法能够快速有效地去除大量外部噪声,并且去噪后的点云数据较好地保留了原始点云特征。In response to the problem of decreased denoising accuracy and efficiency caused by complex data distribution in point cloud collection,this article proposes an anomaly detection method that combines Gaussian distribution model and multivariate normal distribution model.The experimental results show that this method can quickly and effectively remove a large amount of external noise,and the denoised point cloud data retains the original point cloud features well.

关 键 词：异常点检测 高斯分布模型 多元正态分布模型 点云去噪 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]