Comment by Rhett Walker on March 8, 2011 at 7:53am

Interesting concept Denny. So when you say your flying the camera and the UAV follows, your camera angle stays constant while the UAV maintains some kind of balance? you should post a video I'd really like to see what you mean

Comment by Denny Rowland on March 8, 2011 at 9:54am

Actually what I am moving towards is this, The throttle is the normal pressure activated height hold, when yaw is applied the inner axis of the camera mount starts to move. this produces a movement on the cameras LCD screen and the model tries to counter that movement thus following the desired camera angle track. The LCD has a camera chip like the one in your computer mouse that is trained on the center of the LCD image. This data is also used to augment the stabilization of the image within the inner axis system and also form the auto image tracker function. The elevator input controls the cameras pitch movement only and the aileron movement sends the model in a sideways motion, movement forwards is controlled by a manual slider or GPS command. This system enables the camera operator to concentrate on the image acquisition process while the copter follows where it is pointing. There are some complex camera moves that will be autonomously activated using GPS points plus algorithms to blend the circular or curved flight paths. I have quite some work to do yet but the model itself is a major break though in vibration free stabilized flight.

Comment by Sebastian Gralla on March 8, 2011 at 11:13am

would be a big advance if you get it to work like you want it to

 

which hardware do you want to use?

which IMU?

Comment by Matthew Coleman on March 8, 2011 at 12:11pm

I like the camera counterbalance idea.  Maybe you could use the LiPo's as counterbalance weight.

Could it be that those LiPos are a little too warm and cosy in those foam? pockets.  Do they need more breathing space?

Comment by Bill Noyce on March 8, 2011 at 2:20pm

Sounds very cool. My camera mount UAV is still in design stage, but nothing quite as smart as you are developing.

I would love to know where you sourced your damping shocks? I spent an evening online and i found nothing small enough for quadcopter purposes. Any info would be appreciatted.

Comment by Lyn Rees on March 8, 2011 at 5:00pm

Looks to me, "off the cuff" as if the damping shocks are coil over dampers used on R/C cars?

Comment by Lyn Rees on March 8, 2011 at 5:03pm

Nice to see you are developing your airframe to suit a specific application Denny, rather than taking an airframe and adding a camera.

Comment by bGatti on March 8, 2011 at 7:32pm

I believe you can improve shock performance if you move the batteries to the stabilized side.
Also, visual stabilization should be mixed with gyro info to separate alt drift from pitch drift.
Really this is optic flow, you might add 360 degree.

Comment by Denny Rowland on March 8, 2011 at 11:38pm

The model has three batteries and the third one is not shown but yes it is fixed on a slider at the back of the camera mount to aid the final balance of the model and create the balanced beam effect on the mount. Those shocks are only the bump stops in the event of a hard landing. (standard off-road buggy components) they do not attach directly to the copter, but through an airbag platform consisting mostly of bubble wrap and Evo foam. This system can absorb all the vibration you can throw at it. Originally the camera platform was held in a magnetic field, but I felt that it could upset the Mag. so the design was changed to this. I did all of the flight testing with a KK blue board from Minsoo Kim, this was done to provide a bench mark test rig that would allow me to create the ideal aerodynamics platform without the use of Acc. or other stabilization devices. Having exhausted the possibilities of using MK I am now moving to the standard pre-assembled and tested FC from udrones so that the serious development of the electronics can begin. The camera mount has two imus. one for pan and another two axis one that is used for pitch and roll. They use 20deg/sec gyros from Invensense. These are currently integrated into much modified heading lock gyros from various sources that will eventually become a production item. I do however feel that the copters FC will move to 67 deg/sec gyros as the need for 300 is way beyond what is needed for a docile super stable platform, Anyone who has tried to use a 300 deg/sec gyro on the camera mount will have found the noise /oscillation problem that sets in long before you get proper stabilisation. As camera movements beyond 20 deg/sec only produce rubbish footage that can't be used, the 20 deg/sec gyro limits that movement and therefore helps the operator achieve good very stable footage. Post stabilization not spoken here Guys!!!! bGatti are you offering to help? This is a six axis system. the copter being the outer three axis and the mount being the inner high sensitivity part. Normally the outer part ballparks the inner system but here we are much more integrated.

Comment by bGatti on March 9, 2011 at 6:46am

@Denny,

Sounds cool. I'd be thinking in terms of moving all the batteries onto the Camera module.

Optic flow is demanding, one needs either powerful onboards or custom optic flow sensors (mouse sensor is optic flow, but there are (I think) better options posted on this site...

 

Comment

• ‹ Previous
• 1
• 2
• Next ›
• Page