At unexplored terrains or perilous locations, we often wish for a safe and convenient way to know more about place. We dream of autonomous machines, able to give as an accurate image of what to expect, of the place we’ve never seen or too dangerous to go to. Well, thanks to the Kinect, those dreams are becoming a reality as these group of developers from MIT, Washington University and Intel Labs Seattle were able to use a Quadrotor and advances in Visual Odometry in order to create this GPS-Denied Flight contraption with Image Mapping. Flying overhead, having a mounted IMU as it’s central processing unit and the Kinect as the eyes of the machine, the Quadrotor is able to navigate itself around a place, avoiding obstacles and independently guiding itself. In turn, the device can also map the image of the specified place, bringing users a 3d visual image of the place in a remote location.
Here is the description of the project by the developers:
“At MIT, we have developed a real-time visual odometry system that can use a Kinect to provide fast and accurate estimates of a vehicle’s 3D trajectory. This system is based on recent advances in visual odometry research, and combines a number of ideas from the state-of-the-art algorithms. It aligns successive camera frames by matching features across images, and uses the Kinect-derived depth estimates to determine the camera’s motion.”
This project was made possible by the joint efforts of:
Albert Huang, Abraham Bachrach, Garrett Hemann and Nicholas Roy from MIT
Peter Henry, Mike Krainin, and Dieter Fox of the University of Washington
Xiaofeng Ren from the Intel Labs of Seattle.
For more information about the Kinect Quadrotor with Image mapping, visit the developers’ website.