Ender1618
New Member
I have an interest in developing a fully autonomous quadcopter robot specifically for indoor flight.
I am an engineer (what I do in my day job) with experience in robotics, computer vision, navigation/localization, and overall AI.
I was wondering what peoples opinions would be on what quadcopter platform I should attempt to base this project of mine?
So far i am thinking of using an AR Parrot 2, DJI Phantom 3, or a 3DR Solo as the quadcopter.
I currently intend to use a Jetson TK1 as the the onboard computer for control. Also my idea is to use a combination of an on board camera stereo pair, a (or pair of) depth sensor camera(s) (such as a Kinect, or any other OpenNI compatible sensor), a IMU (probably an XSens), along with N ultra sonic rangers.
I know this may seem overly ambitious, but i am familiar with the current research in controlling a quadcopter via on board sensors. I am also adept at developing and expanding upon techniques such as particle filter based localization, visual odometry, SLAM, and Kalman filter based sensor fusion.
The key concept here is that i would like to have this thing be able to fly indoors, and explore its environment.
Any suggestions?
I am an engineer (what I do in my day job) with experience in robotics, computer vision, navigation/localization, and overall AI.
I was wondering what peoples opinions would be on what quadcopter platform I should attempt to base this project of mine?
So far i am thinking of using an AR Parrot 2, DJI Phantom 3, or a 3DR Solo as the quadcopter.
I currently intend to use a Jetson TK1 as the the onboard computer for control. Also my idea is to use a combination of an on board camera stereo pair, a (or pair of) depth sensor camera(s) (such as a Kinect, or any other OpenNI compatible sensor), a IMU (probably an XSens), along with N ultra sonic rangers.
I know this may seem overly ambitious, but i am familiar with the current research in controlling a quadcopter via on board sensors. I am also adept at developing and expanding upon techniques such as particle filter based localization, visual odometry, SLAM, and Kalman filter based sensor fusion.
The key concept here is that i would like to have this thing be able to fly indoors, and explore its environment.
Any suggestions?