IndoorPointCloudViewer

Delft University of Technology
MSc. Geomatics Synthesis (GEO1101) Project
Project 5: 3D Representations for Visual Insight
Download the project report and the introduction video from: http://resolver.tudelft.nl/uuid:a6c4f703-b048-40e3-9661-be00c0fab804
The paper based on this project has been published in LBS2021 conference and it can be downloaded from https://doi.org/10.34726/1786

Team menbers

@Runnan Fu
@Yuzhen Jin
@Zhenyu Liu
@Xenia U. Mainelli
@Theodoros Papakostas
@Linjun Wang

Files structure

geo1101
 ├─ 3Dengine: UE4 project files
 ├─ data: store the input and output files
 │   ├─ pointclouds: store the input and generated point clouds data
 │   ├─ meshes: store the generated meshes data
 │   └─ voxels: store the generated voxels data
 ├─ voxeliser: the module of voxelization, based on the GEO1004 HW01
 ├─ main.cpp: the main workflow of the whole project
 ├─ preprocess.cpp: the workflow of the preprocess
 ├─ preprocess_methods.cpp: functions and data structures used to form the sub_steps in preprocess.cpp
 ├─ basic_methods.cpp: functions used to support the whole project
 ├─ run.py: the trigger of the whole program
 └─ py_methods.py: the functions and data structures written in Python

Instructions

If you only need the final software, you can directly look at section How to run the IndoorPointCloudsViewer (release version).

How to run the pointclouds preprocess program

Install PCL library
For Windows users, it is highly recommended to use Vcpkg to install PCL.
For Mac and Linux users, you can use the official recommended way to install it first. If there are problems in the following steps, you can also use Vcpkg.
For Clion users, Vcpkg is best used in conjunction with the CMake files, as shown in the tutorial here.
Install CGAL library
Make sure that you have installed Vcpkg.
The tutorial for Windows click here.
The tutorial for Mac or Linux click here.
Compile the CPP program
Using Release mode to build the main.cpp, then the build folder will be created.
Set input pointcloud data
Put the input point cloud file in .\data\pointclouds and name as VRR.pcd.
Run run.py
Run the run.py and the output files are in the ./data folder
The meaning of the 4th attribute values in txt files:

0: roof
1: ground
2: architecture part
3: non-architecture part

How to run the IndoorPointCloudsViewer project file in UE4

Install Unreal Engine 4
Install Unreal Engine 4 (UE4) (version 4.26 or later).
Download and unzip the project file
Download IndoorPointcloudsViewer_project.7z from latest release version and unzip it.
Open project file in UE4
The path of UE4 project file: .\IndoorPointcloudsViewer_project\thirdperson.uproject.
Compile and the project
The first run may take a long time.

How to run the IndoorPointCloudsViewer (release version)

Download and unzip the game file
Download IndoorPointCloudViewer.7z from latest release version and unzip it.
Currently the IndoorPointCloudViewer only has Windows version.
Run game file
Open .\IndoorPointcloudsViewer\thirdperson.exe and enjoy it!

Functionality

(A) View the point cloud data in first-person perspective, and the shape of the points is set as circles.
(B) View the point cloud data in first-person perspective, and the shape of the points is set as squares.
(C) View the point cloud in third-person perspective.
(D) View the point cloud data in bird’s eye view, the positions of the avatar and the red target can be identified.
(E) Set the point size as the smallest size.
(F) Set the point size as the biggest size.
(G) View the point cloud in style 1.
(H) View the point cloud in style 2.

Citation

If you want to cite this project in your work, you can use following entry:

Liu, Z., Fu, R., Wang, L., Jin, Y., Papakostas, T., Mainelli, U., & Voûte, R. (2021). Game Engine-based Point Cloud Visualization and Perception for Situation Awareness of Crisis Indoor Environments. In A. Basiri, G. Gartner, & H. Huang (Eds.), LBS 2021: Proceedings of the 16th International Conference on Location Based Services (pp. 183–194). https://doi.org/10.34726/1786

peterliu502/IndoorPointCloudViewer