Say hello to cars without drivers as the Kinect becomes an essential component in developing Autonomous Navigation of Mini Cars. Developers Michael Schweitzer and Michael Himmelsbach are the brains behind this project of using the Kinect’s depth camera capability to determine the range and incoming obstacles of the moving mini car and then send information to a mounted laptop to change the route to avoid collision. They released this video via the youtube account unibwtas which showcases the Kinect and laptop mounted mini car, navigating away from obstacles such as shoe boxes, mini hazard cones, chairs and walls.
A technical description of the technology behind the autonomous navigation is provided in the unibwtas video:
“A planar inverse depth model of the ground surface is verified against the depth images provided by the Kinect sensor to build an occupancy grid of the robot’s environment. The path the robot takes is then determined by a reactive navigation approach called “tentacles”
For more information about this code, you can email the developers by sending inquiries to tas@unibw.de. The code has not been released yet but you can expect to see it here at KinectHacks.com as soon as it is released.
