Today's flight with 'Green Solution'

[Tom Harper]

Photo 1 is a Google Earth overlay of the flight. The segments are color coded - black path is headed for the black triangle. The WP coordinate is the right tip of each triangle. The WP distance for each hit was - black 4, red 3, orange 1, yellow 2, green 4, purple 0.

Used both ailerons this time. The model is very responsive and whips around onto each path. Still a lot of throttle hunting. Just takes a bit more tuning.

The plan was to follow the road around the field. Photo 2 is at the red WP, photo 3 is half way between the orange and yellow triangles. Photo 4 is at the green WP.

With this slow camera there is no overlap. Resolution is good, even with the low setting. Will double the altitude. May try some trail photos tomorrow..

Good job Dean!!

[CenTexFlyer]

Good imagery Tom

Were you using the throttle reduction on picture?

Gene

[Tom Harper]

Gene,

No throttle reduction. Camera is an Olympus C-50, 1/1000 sec, ISO 340. By slow, I was referring to the frame rate of 4 seconds per image.

This looks adequate for photographing historic wagon trails.

Tom

[CenTexFlyer]

Indeed it is! We've been using the Lumix and Canons and vibration has cropped it's ugly head again. Seems the OIS is fighting us to produce blurry shots even at shutter speeds over 1/500 second.

[Gary Mortimer]

Well done Tom, I am utterly jealous.

We stitch images and as long as you go back and forth over the area at the same altitude, you can end up with pretty good results.

I will attach an example of a dig in Falkirk.

Currently we are normally taking one day to get the shots and two days post production.

Cheers

G

[Tom Harper]

Gary,

Good record - if that is stitched, it's really tight!

I will try to get overlap tomorrow so I can try stitching.

Tom

[Gary Mortimer]

Just get stacks of shots from the same altitude, I guess just creating lots of paths that cross each other, maybe fly east west down a square and then north south up and down same square.

I think that one was 20 odd images.

You also have to do lots of hand filtering for best results.

As I say takes lots of post production time, I look forward to the results!!

The cropcam takes its images at 2100' I think, now thats very naughty of course.

Of course the lower you are the better resolution you will get, no point in recreating google earth res images.

Cheers

G

[CenTexFlyer]

Gary,

That's an oustanding mosaic of the Falkirk dig. I have to agree with you that a lower altitude and higher resolution outshines just about anything else out there.

Good work both of you!

Gene
(Gunn Clan - Aut Pax Aut Bellum)

[Gary Mortimer]

Thankyou Gene

My mate Mike is the whizz with the stitching, I don't have the patience he has with it!

Heres a higher level stitch of a lost settlement.

The vertical camera mount being a hole cut through a flying wing.

Then as mentioned before fly around alot at the same altitude and chuck the resulting 3 or 400 images into a stitching program.

And wait!

Its the way forward for your trail shots and I will employ it on another task in South Africa later in the year when I get home.

I have a stack of tasks waiting for an Atto here.

G

[Tom Harper]

I spent a lot of time with the log file and the images - finaly realized that I had about 20% overlap. So, I loaded the stitch program and - how about that - it got stitched. The result has great resolution (about 111 meg), but I had to squash it down to load it here. Attached is a sequence of 8 images.

I see what you mean. Looking closely at the stitch lines I can see that the match is crude. If I size and align the images, the stitcher will do a better job.

Tom

[CenTexFlyer]

Great job, Tom! I'd say your well on your way to some really polished results!

[Gary Mortimer]

Its a mission to size images Tom, rather use the attopilot to hold the alt nicely for you and get lots and lots of images.

Then hand filter them for sharpness and levelosity (made up word)

Once you've done that, crunch them. (not credit)

Post production takes much longer than grabbing the images.

Cheers

G

[Tom Harper]

Tried a field test this AM. The wind was from the South and gusting 6 to 10 MPH. That meant a cross wind take off and landing if we used the county road. So we tried a couple of other locations and settled on a hand launch (cliff toss) to the South from a point where we could observe the entire flight. The plan was for a 430 ft altitude at 45 KMH.

As you can see on the plot the hand launch went well (gray), we switched to auto after about 10 seconds (red). The model ran with the wind then took up and held a heading of about 60 degrees which hit the first WP within 2 meters (yellow) (the triangle is a bit off because I did not lock the wps to ground for the plot).

It made a smooth 180 and came around on 240. At a minute and a half into the flight something happened. It left the heading and went nutso, eventually falling into a vertical spin. Carl hit the throttle and saved the plane. It was a bumpy ride back. Problems seemed to be associated with reducing the throttle. He brought it right back to the edge of the cliff but was a meter short on landing. There goes the pitot tube again.

Probably need to hold an airspeed above 50 kmh and further reduce the throttle gain - it's still hunting.

Maybe early tomorrow.

[boyisabird]

Then as mentioned before fly around alot at the same altitude and chuck the resulting 3 or 400 images into a stitching program.

And wait!

G

Gary-
Which stitching program do you use?
Thanks Dave

[dmgoedde]

Tried a field test this AM. The wind was from the South and gusting 6 to 10 MPH. That meant a cross wind take off and landing if we used the county road. So we tried a couple of other locations and settled on a hand launch (cliff toss) to the South from a point where we could observe the entire flight. The plan was for a 430 ft altitude at 45 KMH.

As you can see on the plot the hand launch went well (gray), we switched to auto after about 10 seconds (red). The model ran with the wind then took up and held a heading of about 60 degrees which hit the first WP within 2 meters (yellow) (the triangle is a bit off because I did not lock the wps to ground for the plot).

It made a smooth 180 and came around on 240. At a minute and a half into the flight something happened. It left the heading and went nutso, eventually falling into a vertical spin. Carl hit the throttle and saved the plane. It was a bumpy ride back. Problems seemed to be associated with reducing the throttle. He brought it right back to the edge of the cliff but was a meter short on landing. There goes the pitot tube again.

Probably need to hold an airspeed above 50 kmh and further reduce the throttle gain - it's still hunting.

Maybe early tomorrow.Tom - You really need my newest firmware. It won't let the ground speed go below 15 km/h no matter how on target the airspeed is; if the ground speed starts to sag then extra power is added. Of course this only works if the plane has the power to make it work.

Check your e-mail.

[Tom Harper]

Dean,

Thanks!

I see it has 'Auto Lock Distance' - that is a feature I was looking for. Good thing we were not using it yesterday though.

Is the distance in meters or Km?

Tom