3D Map Update System with LIDAR Trace and HD Map Integration

This project enables 3D map updates using LIDAR traces and a high-definition (HD) map of an environment. The system is divided into three main components:

1. Visualizer - Visualizes LIDAR trace and map updates.
2. Client-Server Model - Facilitates storage, updates, and downloading of the base map.
3. CMake Project - Processes LIDAR trace data to detect changes using a custom ray-tracing algorithm.

Project Structure

• visualizer.py: Visualizes the LIDAR data and map updates.
• map_cloud_server.py & map_cloud_client.py: Components of the client-server model for base map updates.
• CMake project for ray-tracing algorithm-based trace processing.

Setup

Visualizer Setup

Create and activate the environment:

conda create --name open3d-env python=3.6 # Create the environment
conda activate open3d-env # Activate the environment
pip install open3d==0.15.1 # Install Open3D

After configuring your file paths, run:

python visualizer.py

You can use , . to go back or forward one frame respectively, depending on your step size. You can also pause, move around, ctrl+click, right click, and press R to reverse the player.

Client-Server Model Setup

The client-server model allows storing, updating, and downloading changes to the base map.

python map_cloud_server.py # Run the server
python map_cloud_client.py # Run the client after setting the input trace path

CMake Project Setup

The CMake project utilizes a previously captured LIDAR trace for detecting map updates. It features a custom ray-tracing algorithm for efficient change detection.

1. Ensure CMake is installed on your system.
2. Configure the project, build and run:

cmake 
make
./3dMapUpdates

3. Run the executable to process the LIDAR trace.

---

Contributions

Feel free to contribute to the project by forking the repository, creating new branches, and submitting pull requests!

Installation for Linux (works on school machines)

1. You might want to update apt apt update
2. Install cmake apt install cmake
3. If you don't already have git, get it: apt install git
4. Git clone vcpkg git clone https://github.com/microsoft/vcpkg.git
5. In vcpkg directory cd vcpkg, run the file ./bootstrap-vcpkg.sh
6. Integrate install: vcpkg integrate install
7. Install PCL with visualization module - vcpkg install pcl[visualization] (This can take 30 minutes or so)
8. Git clone the repo git clone https://github.com/samsitGit/3dMapUpdates
9. Go to the project cd 3dMapUpdates then configure the project with CMake: cmake
10. You may want to go into your 3dMapUpdates.cpp to change file paths for your ego and a complete trace unless you are on Irfan's computer.
11. Compile the program to get an executable: make
12. Run the file ./3dMapUpdates

Installation for Windows

1. Follow this to set up and install CMake and vcpkg: https://learn.microsoft.com/en-us/vcpkg/get_started/get-started-vs?pivots=shell-cmd
2. Run this command to set up vcpkg related libraries: vcpkg install
3. Install PCL with vcpkg - vcpkg install pcl
4. Add vcpkg to your PATH in environmental variables.
5. Add vcpkg as VCPKG_ROOT to your environmental variables.
6. Modify your CMakePresets.json to set all "DCMAKE_TOOLCHAIN_FILE" (there are multiple) to vcpkg.cmake located in scripts > buildsystems folder where you installed vcpkg.
7. Now you can open this diretory in Visual Studio (install it if you haven't), and it will configure it.