jjihwan / AR_stereo_vision

2023 Spring SNU Computer Vision Project

Display Virtual Cubes on My Desk

: 3D Reconstruction and AR via Stereo Vision on Videos

2023 Spring SNU Computer Vision Project by

Jihwan Kim [e-mail]
Hojune Kim [e-mail]

How to use

git clone
cd 2023S_SNU_CV_Project
conda env create -f "environment.yaml"
conda activate cvproject
cd liegroups
pip install .
cd ..

If conda doesn't work, try :

conda create -n "happy"
conda activate happy
pip install numpy opencv-python matplotlib
cd liegroups
pip install .
cd ..

and then,

python main.py

Note that you should click one more point on second plot

Usage

To verify camera calibration,

python main.py -c

To choose other video,

python main.py -v="./core/data/short.MOV"

To designate output video path,

python main.py -o="./core/results/output.mp4"

To see optical flow processes for each frame,

python main.py -d

Then you can find results in
./core/results/VIDEO_NAME

Rule

all files should be located in ./core without main.py
all images, data files should be located in ./core/data
all results should be located in ./core/results