Skip to content

This is an cross platform 2D radar labeling tool for NCU MRL Data Science course

Notifications You must be signed in to change notification settings

NcuMathRoboticsLab/MRL_LabelTool

Repository files navigation

MRL_LabelTool

(The video was speeking in Chinese)
How to use: https://www.youtube.com/watch?v=m87yXTsJ6vg
Build Tutorial: https://www.youtube.com/watch?v=PMmqYjvkA2Y

Noticed that the name of the project folder(root folder) must be MRL_LabelTool

This is an label tool for MRL Project, it include two windows, which can help you label the data and check the Model.

This project use C++20, Eigen3, ImGUI, ImPlot, ImGuiFileDialog to build, it was all included in 3rdparty folder.
This project also supports C++17.

On Windows10, it builed by Vscode and CMake, you can use Visual Studio, too.
On Linux you can just build by CMake.

Course Notice

There is an Adaboost with logistic regression example in this project.
In this courese, you must implement your own model, otherwise you won't pass the course.

If you wanna implement your model in this project directly.
To make your own feature, you can modify the "include/make_feature.h" file.
To make you own Model, you can add your model in the model folder, notice that there must be two member functions, fit and predict in your model.

Or you can implement your model in your own project, then embed it into the GUI tool.

You can refer to the show_simulation_window.cpp and show_label_window.cpp to embed it.
If you have any question about the GUI tool, please contact the project maintainer.
You can find them in MRL. Or you can ask the TA for the contact details of the maintainer.

Dependencies

  • Eigen3
  • ImGUI
  • ImPlot
  • ImGuiFileDialog
  • OpenGL
    On Windows, you can download it in Visual Studio.
  • glfw
    On Windows, you have to build it manually.

Linux

just clone it and build it.

git clone https://github.com/Mes0903/MRL_LabelTool.git
cd MRL_LabelTool
mkdir build && cd build
cmake ..
cmake --build .

Windows

You can check the Demo Video to help you build it, it's a little complicated.

glfw

There is a lib file glfw3.lib in the 3rdparty/windows/lib, you should build it by yourself. (Since the lib file on the github is depends on my environment.)

  1. download the source code from glfw.
  2. make a build folder and get into it (You can do it by Vscode)
  3. cmake ..
  4. Open the GLFW.sln, build the project by Visual Studio.
  5. There would be a lib file called glfw3.lib in Debug folder, delete my glfw3.lib and copy your glfw3.lib into 3rdparty/windows/lib.
  6. Also, there is a folder called GLFW in the glfw 3rdparty/windows, delete my GLFW and copy your GLFW into 3rdparty/windows.

glad

  1. There is a CMakeLists.txt in the 3rdparty/windows/glad, copy it to somewhere first, then delete the 3rdparty/windows folder in the project.

  2. Download the glad correspond to your Opengl Version on the glad loader website. Then move the glad folder into 3rdparty/windows.

Contact

If you find any bugs, PR and issues are wellcome.

Or you can contact TA or Mes([email protected]).

About

This is an cross platform 2D radar labeling tool for NCU MRL Data Science course

Topics

Resources

Stars

Watchers

Forks

Packages

No packages published