Natural Interaction via REST (NIREST)

NIREST uses the OpenNI2-FreenectDriver from libfreenect of the OpenKinect Project, OpenNI, NiTE, and Linked Data-Fu

This application provides body tracking for Kinect v1 as RDF read-only REST resources or via a Linked Data-Fu API.

Requirements

• Linux
• Connected Microsoft Kinect v1
• Java 8
• Docker (optional for running the NIREST Docker container)

Setup

All required commands are provides as shell scripts, which may be read for further investigation.

Docker

NIREST Docker images are automatically build at Docker Hub.

• Based on latest Ubuntu image
• Installation of all dependencies
• Cloning of NIREST and libfreenect repositories
• Building and installation of libfreenect, including OpenNI2-FreenectDriver
• Installation of OpenNI and NiTE middleware
• Configuration of OpenNI and NiTE

Docker Image

Pull the latest docker image.

./docker-pull.sh

Docker Helpers

Run the docker image and start the NIREST server, which will be available at "http://localhost:8888".

./docker-run-server.sh

Run the docker image and start NIREST with integrated LD-Fu API.

./docker-run-ldfu.sh

Run the docker image and start a Bash session for manual intervention.

./docker-run-shell.sh

Docker Build

For manual local build of the docker image, clone the GitHub repository of NIREST and build the Docker image as usual.

./docker-build.sh

Gradle

Run NIREST server directly via Gradle using pre-compiled OpenNI2-FreenectDriver.

./gradle-run-server.sh

Run NIREST with integrated LD-Fu API directly via Gradle using pre-compiled OpenNI2-FreenectDriver.

./gradle-run-ldfu.sh