This is a project for "Advanced Topics in 3D Computer Vision" Praktikum course, offered by TUM in Winter Semester 2019. In this project, we have implemented a pipline to do a 3D recunstrion of a scene, using a depth camera, segment the objects and then use the data to do augmentations.
These instructions will help you running the code on your local machine.
- One of the libraries we use is Open3D, which does not support Python 3.7. We recommend running the code using a Python Enviroment.
- For Mac users, the SDK of the Intel® RealSense (the RGBD camera which we use) is not available, so you have to build the lib files from source and add them to the 'pyrealsenseformac' folder.
To run the whole pipline with UI, go to the source directory and run main.py.
python main.py