Skip to content

mohith-sakthivel/deep_hand-eye_calib

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

68 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Hand-Eye Calibration in Unstructured Environments

In this project, we present a deep learning based method to perform hand-eye calibration in the wild. Our method does not require any object of known geometry to be present in the scene. It estimates the hand-eye calibration parameters by directly making sense of the visual features present in the images taken from different view-points.

For more details, visit the project website.

Note

  • The quaternion convention followed in this repository is (w, x, y, z)

Setup

To setup the environment

git clone https://github.com/mohith-sakthivel/deep_hand-eye_calib.git
cd deep_hand-eye_calib

conda env create -f environment.yaml
conda activate hand-eye

To download and setup the dataset

wget http://roboimagedata2.compute.dtu.dk/data/MVS/Rectified.zip
unzip Rectified.zip -d data/DTU_MVS_2014

mv temp/camera_pose.json data/DTU_MVS_2014

To train the model

python -m deep_hand_eye.train

Releases

No releases published

Packages

No packages published

Languages