基于八叉树编码的点云压缩研究与实现

发布时间：2023-04-02 07:54

　　点云压缩是目前计算机视觉、机器人、以及虚拟现实和增强现实领域的热点话题,它正成为发展最快的技术之一,但在将其有效地应用于各种实际场景之前,仍需要攻克一些难题。当在各种应用中进行实现时,点云数据的大小仍是一个主要问题。本文基于点云压缩的各种现有技术,主要比较了两种点云压缩技术的性能,即基于双缓存八叉树点云编码和基于单八叉树的点云编码,并对这两种技术展开了详细分析和讨论。此外。论文提出一种改进的点云压缩八叉树算法,称为单八叉树静态点云压缩算法,此算法的提出受到双缓存八叉树编码算法(Kammerl,2012)的启发。本文所提出的单八叉树静态点云压缩算法使用与双缓存八叉树压缩类似的压缩模型,遵循网络上点云压缩和解压缩的整个生命周期。它通过使用八叉树数据结构在空间上分解点云来进行初始化,然后使用二进制流对该结构进行比特掩蔽和序列化以表示用于编码和解码点信息的八叉树结构。此外,它使用点位置编码来编码点的附加信息,例如,颜色,NORMALS等。仿真结果显示,本文所提出的八叉树算法对于静态点云具有比双缓存八叉树算法更好的压缩性能。相比于双缓存八叉树算法,虽然所提方法的视觉性能没有改善,但压缩比明显提高...

【文章页数】：60 页

【学位级别】：硕士

【文章目录】：
ACKNOWLEDGEMENT
ABSTRACT
摘要
List of Abbreviations
Chapter 1:Introduction
    1.1 Overview
    1.2 Motivation
    1.3 Problem Statement
    1.4 Thesis Outline
Chapter 2:Key Technologies
    2.1 Point Cloud Compression
    2.2 Compression Techniques
    2.3 Point Cloud Library (PCL) Properties
        2.3.1 Filtration
        2.3.2 Features
    2.4 Data Structures for Point Cloud Compression
    Summary
Chapter 3:Related Works
    3.1 Achieving Real-Time Compression using Double Buffer Octree Technique
    3.2 Double Buffer Octree
    3.3 Conventional Image Compression Technique for data transmission for Point Clouds
    3.4 XML based Compression Scheme for dense point cloud streaming
    3.5 Efficient Processing of Large 3d point cloud
    3.6 Octree-based Algorithm for Scattered Point Clouds
    3.7 Combination of Octree and Quadtree for Point Cloud Compression
    3.8 Usage of Octree Data Structure
    Summary
Chapter 4:Single Octree-based Static Point Cloud Compression
    4.1 Octree Based Encoding for Point Cloud Compression
    4.2 Significance of Outliers removal for Octree and Compression
    Summary
Chapter 5:Simulation Results
    5.1 Tools and Technologies
    5.2 Compression Results
    5.3 Double Buffer Vs Single Octree
    Summary
Chapter 6:Conclusion and Future Work
    6.1 Conclusion
    6.2 Future Work
Reference



本文编号：3778851