AzureKinect is a program for viewing and recording of data from an Azure Kinect camera. It can can view/record the direct image data from the camera as well as perform body tracking calculations and visualisation.
- View Azure Kinect camera depth image real-time output
- View Azure Kinect camera 4k colour image real-time output
- View Azure Kinect camera Infra-red image real-time output
- Use GPU accelerated AI based body tracking
- View the calculated body tracking region of interest and determined skeletal positions
- Record the calculated body tracking data to a csv file
- Record any/all of the Azure Kinect cameras to h264 in real-time
- Use CPU or GPU accelerated h264 encoding
- 60fps visualisation, recording and processing
- Microsoft Azure Kinect camera
- NVIDIA GPU RTX2070 or better
- Windows 10 April 2018 (x64) or newer
- 4+ core Skylake or better CPU
- 8GB+ memory
- Recording simultaneous streams in real-time may require additional CPU resources
Compilation uses Visual Studio. To build the program simply open the supplied project.
The following dependencies are required in order to build:
- Visual Studio 2019 or newer
- Qt5 SDK
- Qt Visual Studio Tools addon
- Nuget (used to install Azure Kinect SDK, glm and FFmpeg)