KinZ for Python A library for Azure Kinect

At the time of this writing, there are no official python bindings for Kinect for Azure.
This library allows to use Azure Kinect directly in Python.

Installation:

Install the Azure Kinect SDK as described here where it says Microsoft installer. For Windows, Download the .exe and follow the steps. For Ubuntu use the sudo apt install commands shown in the same page.
For Body tracking functionality (optional), you need an NVIDIA GPU and install CUDA. Download from here.
For Body tracking functionality (optional), install the Azure Kinect Body Tracking SDK. For Windows, download the msi installer from here. For Ubuntu simply run the sudo apt install command provided in the webpage.
Before compiling the code for Matlab, make sure the Kinect works correctly using the viewers provided my Microsoft, e.g. C:\Program Files\Azure Kinect SDK v1.4.1\tools\k4aviewer.exe and C:\Program Files\Azure Kinect Body Tracking SDK\tools\k4abt_simple_3d_viewer.exe. In Linux just type k4aviewer or k4abt_simple_3d_viewer in the terminal.
Install the Python requirements (see environment.yaml). To create a fully functional conda environment run:

conda env create --file environment.yaml

Install the library with:

pip install .

To install KinZ with body tracking functionality run:

CFLAGS="-l:libk4abt.so -DBODY" pip install .

Demos

Inside demos directory, you'll find demos showing all the features of the library.
Currently, there are only 6 demos:

cameras_sensors_demo: shows how to get color, depth IR, and IMU sensors.
calibration-demo.py: shows how to extract camera calibration values.
map_depth_to_color_and_3d_demo.py: shows how to map sparse points from depth to color and from depth to 3D.
map_color_to_depth_and_3d_demo.py: shows how to map sparse points from color to depth and from color to 3D.
pointcloud_demo: shows how to get the colored pointcloud and visualize it with open3D.
body_tracking_demo: shows how to get body tracking information and visualize it.

Streaming Speed

The streaming speeds in Python are the following:

Resolution	FPS Unaligned	FPS Aligned Depth2Color
1280 x 720	30	30
1920 x 1080	30	30
2560 x 1440	30	22
2048 x 1536	30	30
3840 x 2160	30	16
4096 x 3072	15	9

Uninstall:

pip uninstall kinz

Coming soon ...

Hand pose estimation

Color image keypoints

Depth image keypoints

For more details checkout the wiki pages.

Citation

@article{terven_cordova_kinz,  
title = {KinZ an azure kinect toolkit for Python and Matlab},  
journal = {Science of Computer Programming},  
pages = {102702},  
year = {2021},  
author = {Juan R. Terven and Diana M. Córdova-Esparza},  
keywords = {Azure Kinect, Python, Matlab}}

Link to paper

jrterven / KinZ-Python