基于微软Xbox Kinect传感器的三维场景重建方法研究
本文通过相对廉价的Kinect设备分别从不同视角采集了场景中的深度点云数据，利用PCL点云库对深度点云数据进行了滤波采样，并利用PCL库中封装的ICP迭代最近点算法进行了配准工作，得到了一幅三文场景重建图。尽管由于设备精度和算法缺陷的原因，最后经配准的三文场景图有一定的重影并且一些细节不够精确，但是三文场景重建的目的基本达到，可以满足简单的实验需求。
毕业论文关键词 Kinect 三文场景重建 PCL点云库 ICP
毕业设计说明书外文摘要
Title The Study Of Three-dimensional Scene Reconstruction method based on Microsoft Xbox Kinect sensor
Abstract
Three-dimensional scene reconstruction is to study how to obtain three-dimensional information of the scene with two-dimensional information of the scene which has been collected. It relates to computer graphics and image processing technology. Compared to the two-dimensional information, three-dimensional information giving more realistic to us. Therefore ，three-dimensional scene reconstruction has high research value. This paper collected the depth point cloud data of the scene from different perspectives by relatively cheap Kinect sensor. Filtered and sampled the depth point cloud data by Point Cloud Library. And using ICP （iterative closest point） algorithm encapsulated by PCL library to carry out registration. Finally this paper obtained a three-dimensional scene reconstruction Figure. Although due to the precision of equipment and the defect of algorithm, there are some ghosting in three-dimensional scene graph which has been registered. But the purpose of the three-dimensional scene reconstruction basically reached, and it can meet the needs of simple experiment.
Keywords Kinect 3D reconstruction PCL ICP
目 录
1 绪论 1
1.1 论文的研究背景 1
1.2 国内外研究现状 1
1.2.1 光度立体学法 1
1.2.2 运动图像序列法 1
1.2.3 立体视觉法 2
1.2.4 三种三文重建方法的比较 2
1.3 基于图像三文重建的基本步骤 3
1.4 本文主要研究内容及组织结构 4
2 Kinect技术原理 5
2.1 Kinect简介 5
2.2 Kinect深度传感器工作原理 6
2.3 Kinect的应用 7
2.4 OpenNI简介 8
2.5 点云库PCL简介 9
3 点云数据的获取与存储 11
3.1 参考坐标系 11
3.1.1绝对坐标 11
3.1.2 相对坐标 11
3.2生成数据前的准备 11
3.2.1错误信息返回 11
3.2.2上下文对象初始化 12
3.2.3 创建生产节点 12
3.2.4 设置图输出模式 12
3.2.5 图像位置的矫正 13
3.3 点云数据的获取 13
3.3.1 建立点云数据结构 13
3.3.2 产生数据和更新数据 14
3.3.3 获得数据并进行坐标转换 14
3.4 存储点云数据 16
3.5 停止产生数据 17
上一篇：鱼眼镜头图像畸变数字校正方法
下一篇：多频信号压缩采样与高分辨估计