3d点云压缩系统的研究.docxVIP

本科毕业论文(设计） 题 目： 3D 点云压缩系统的研究 姓 名： 巴·孟克吉尔格力 学 号： 201301121001 学 院： 信息科学与工程学院 专 业： 通信工程 指导教师： 元辉 摘 要 近年来，随着三维扫描设备的快速发展，三维数据的扫描成本大大降低，采 集三维数据变得越来越容易、快捷。点云作为一种有效、高质量的三维数据获取 及存储方式逐渐被广泛接受并使用,点云数据的处理技术也愈发成熟。本文将重 点研究三维点云数据的压缩系统，对大量点云数据进行压缩精简的预处理，主要 通过构建一个编码/译码器来实现压缩/解压缩大量密集点云数据。在该压缩系统 中，我选择基于八叉树数据结构的压缩方法，通过预测帧编码实现相关功能。此 外对于三维可视化数据（图像、视频等），其他相关属性，如颜色、深度也相当 重要，因此我将选择颜色属性在该压缩系统中进行有效编码以实现对彩色图像的 高质量压缩/解压缩。文中采用的开发工具为开源 C++编程库点云（Point Cloud Library，PCL）库配合 Microsoft Visual Studio 以及一种跨平台的编译工具 CMake，以 C++作为主要编程语言实现对大量点云数据的压缩/解压缩功能。 关键词：点云压缩；点云库；八叉树 Abstract With the rapid development of 3D scanning devices, it is becoming easier and convenient to gain 3D data at a low cost.At the same time,Point Cloud ,as an effective and high-quality way to get 3D visual data, is widely accepted and used which lead to the improvement of technologies of point cloud data processing.This paper mainly focuses on the compression system of 3D point cloud which aims to implement a encoder/decoder to compress/decompress huge amounts of point cloud data.In this system,I choose the method based on octree algorithm and use PCL as developing kit.Prediction frames are coded to improve the compression rate.Moreover, this paper represents the additional attribute,such as color, which is of great importance of visual data like image and video,can be efficiently coded in this system.The developing kit for this system is based on open source C++ programming Library PCL ,Microsoft Visual Studio(MSVC) and a cross-platform free and open-source software CMake to implement the compression and decompression of large amounts of point cloud data. Keywords：Point Cloud Compression; PCL; Octree 目 录 摘要.................................................. Abstract................................................ 第一章 绪论................................................1 1.1 选题背景与研究意义....................................1 1.2 三维点云数据压缩技术简介...............................1 1.3 开发工具介绍..........................................2 1.3.1 点云库....