Comment by Martin Poller on April 18, 2012 at 4:36am

Hi James,

Would you mind answering some more questions please.

1. Are you still using APM1 or have you move to APM2 now, and if you are still using APM1 what version of the Arduplane software are you using? Is it now performing how you would like?

2. How doe the folding prop work with the plane being a pusher, do the blades still fold towards the back of the aircraft?

3. Have you got a working gimbal now, did you base it on the one that you have linked to on youtube.

4. Are you using two cameras, one IR and one RGB? if so are you synchronising them?

Thanks

Martin

Comment by James on April 18, 2012 at 4:52am

Hi Martin,

1. I just finished my APM2 Setup, but never tried it, all my work, done with APM1 and updated to every stable release, the performance was great every time. 

2. Fold prop saved me about 3-4Amps so its working great, plus the looks

3. i use just a roll gimbal, basically because its all i need and its easier to setup & fit to fuselage

4. Iam experimenting with the following setups: A. 1modified Nir,G,B Cam for vegetation studies. B.1 RGB P&S for aerial photography. C . 1 RGB And 1 IR, NIR Combined for super results but more home work for combining.

Here is a video just showing of  aircraft stability:

http://youtu.be/6p_jIDYnVBU

And a video with my ground station:

http://youtu.be/TBTf--oOsl4

Comment by James on April 18, 2012 at 4:55am

some photos

Comment by James on April 18, 2012 at 4:55am

Comment by Martin Poller on April 18, 2012 at 11:00am

Hi James,

Thanks for the answers,

Can you tell me are you using two propellors?

Comment by James on April 19, 2012 at 2:37am

good morning, your are welcome,

Basicaly 100 +2 propellors during tests :)

ΑPC E (from 11-5.5 to 9-6 ) Style are more efficent but with more noise and drag, Folding are a little less efficient  in thrust but help gliding a lot..

Here is an orthophoto- pseudocolor from G, B, NIR (15cm Resolution) From modified a canon ixus 100is: 

Comment by James on April 19, 2012 at 2:41am

And NDVI analysis of the area:

Comment by Willem Brussow on July 30, 2012 at 1:05pm

Hi James,

I am on a engineering co-op in Canada at the moment and we are just starting to do NDVI. How did you stitch these images? Will Pix4D be able to do it? Also, did you process the NDVI images before stitching them, or did you create a big geotiff of the area for bor RGB and NIR and then do the NDVI "post-rectifying"?

Much Appreciated,

Willem

Comment by James on July 31, 2012 at 2:02am

Hi Willem, nice to meet you

There are many ways to stitch images (pix4D, Microsoft ICE, Erdas imagine tools, ArcGIS tools). For the images above, i used the most painfull way- because its a vinyard- and i wanted seemless image and perfect accurasy:

-I rough geoferenced every image seperatly on ArcGIS using my logs and simple home made routine,

-then i updated georeference for each one using ground control points dropping rms to 2.55cm! ,

-then using the raster processing tools i made a mosaic,

- finaly using a 4m DEM, from our cadastral agency, i ortho-corrected the hole mosaic.

i made the same for both nir images and RGB images,  then again using raster prossesing ArcGIS tools i used composite bands.

Next enhancements: apply  a lens error filter to every captured image, also calculate NIR 8bit scaling Vs Solar radiation in order to compare places and captured images 

All these seems like a lot of work, but if you use a powerfull GIS soft (many opensource are equally powerful), you can create routines and once you do it, you can repeat you work for another site in seconds.

I hope i help'd 

James

Comment by eduardo on October 9, 2012 at 10:06am

wow ... 40minutes with 1x 2600mha 4s battery ? what wind speed ?

I never get more than 30 minutes using 8000mha on my 2.2m flywing on 8m/s wind.

Comment

• ‹ Previous
• 1
• 2
• 3
• Next ›
• Page