Computer vision project focused on augmented reality using marker detector and pose estimation.
This repository contains the source code for Assignment 3 of Geometric and 3D Computer Vision course. It aims at performing an augmented reality by using marker point coordinates detected in the previous assignment, repository polygonal-marker-detector, to compute marker pose and project a virtual cube on top of a rotating object using augmented reality techniques.
The main.ipynb jupyter notebook contains the code to track objects in a video, extract camera pose, and overlay a 3D cube on the object in each frame of the video. The intrinsic parameters K and the distortion coefficients are already provided:
- K: [[1.66750771e+03 0.00000000e+00 9.54599045e+02] [0.00000000e+00 1.66972683e+03 5.27926123e+02] [0.00000000e+00 0.00000000e+00 1.00000000e+00]]
- distortion: [ 1.16577217e-01 -9.28944623e-02 7.15149511e-05 -1.80025974e-03 -1.24761932e-01]
Before of running the main notebook, there are some requirements:
- Python 3.
- The following python modules:
pip install opencv-python pip install numpy
- The dataset folder, which can be downloaded here.
The main notebook can be run from any IDE which supports Jupyter Notebooks.
git clone https://github.com/elaaj/augmented-reality