Survey v1 UI: Reducing options

[bbasso]

In the Advanced Options, Grid Options tab, Overshoot should just be one option, and the same on each side of the polygon.

Overlap and Sidelap can be the same number as well, as long as they are area-wise, not length-wise. ie. 60% overlap means 60% of the area is the same, regardless of if it's on the side or the top of the sensor footprint.

[iskess]

I'm using all of these options and would not want to see them go. If I have a strong wind parallel to the flight lines I want a larger overshoot on the downwind side than the upwind side. I adjust my overlap according to the speed of my camera, and my sidelap might be different depending on turbulence, whether I'm using my roll gimbal, or many other factors.

[chuckd1974]

Yes I find all of these options very valuable especially when having to report these parameters for academic papers while I am trying to graduate. Here in Hawaii we have very dynamic wind conditions that we have to adjust the mission plan to accommodate.

[bbasso]

Ok, I can see a reason to have two overshoots. Two overlaps don't make sense, as stitching software only cares about the area-wise overlap, in any direction.

[ReMapHI]

The ability to adjust sidelap is still important. If you only had overlap but no sidelap, you would still have problems matching images on adjacent flight lines.

[bbasso]

I think when feature-based stitching software says it requires a specific overlap, they're really trying to place a lower bound on the number of tie points you will be able to find in adjacent images. So in that sense all that matters is the area overlap in any orientation, even if it's just on a corner and not a side. Having worked with the OpenCV toolbox, I can say for sure that the % area overlap is all that matters, laterally, longitudinally, diagonally, whatever...

So 50% "overlap" would enforce 50% lateral and longitudinal overlap as a percentage of area.

[ReMapHI]

I still think it's important to adjust endlap and sidelap separately. I understand the desire to keep things simple, but I believe that Mission Planner could be further improved by allowing the user to adjust individual parameters according to the local conditions (wind/terrain), aircraft performance, camera capabilities, and image processing requirements. Commercial flight planning software packages offer much more control over these parameters.

We often fly in strong winds over complex terrain, which poses challenges for preventing gaps in between flightlines. The ability to adjust sidelap is helpful to account for the effects of strong winds (drift, crab and roll) as well as variations in flying height and terrain.

[chuckd1974]

Brandon, Thanks so much for the amazing and useful tool in MP. It would be nice to have the control option for both lateral and longitudinal overlap. In Hawaii, we have so much variation in wind and topography that we simply need the sidelap to assure total coverage when the bird gets blown off track few several meters. Also some of our research involves characterizing topography in 3 dimensions. The clustering multview stereopsis step in the processing is dependent on the feature matches generated during the scale invariant feature transform. SIFT works best, according to the literature, when the feature is visible in as many camera locations as possible. I would grateful if the option was left in MP.

Mahalo,

CHuck

On Tue, Jul 29, 2014 at 8:36 AM, ReMapHI notifications@github.com wrote:

I still think it's important to adjust endlap and sidelap separately. I understand the desire to keep things simple, but I believe that Mission Planner could be further improved by allowing the user to adjust individual parameters according to the local conditions (wind/terrain), aircraft performance, camera capabilities, and image processing requirements. Commercial flight planning software packages offer much more control over these parameters.

We often fly in strong winds over complex terrain, which poses challenges for preventing gaps in between flightlines. The ability to adjust sidelap is helpful to account for the effects of strong winds (drift, crab and roll) as well as variations in flying height and terrain.

— Reply to this email directly or view it on GitHub https://github.com/diydrones/MissionPlanner/issues/542#issuecomment-50518942 .

Charles Devaney Masters Candidate Department of Geography University of Hawaii at Manoa

[bbasso]

Interesting--I guess I can see the argument: cross-track errors are more likely than along-track errors, so you want more sidelap to guarantee good coverage. I'm convinced--lets keep it in there!

[chuckd1974]

Thank you Brandon! The awesome work you do is greatly appreciated!

Chuck

On Tue, Jul 29, 2014 at 11:22 AM, Brandon Basso notifications@github.com wrote:

Interesting--I guess I can see the argument: cross-track errors are more likely than along-track errors, so you want more sidelap to guarantee good coverage. I'm convinced--lets keep it in there!

— Reply to this email directly or view it on GitHub https://github.com/diydrones/MissionPlanner/issues/542#issuecomment-50540669 .

Charles Devaney Masters Candidate Department of Geography University of Hawaii at Manoa

[jmachuca77]

I agree with keeping the options I was actually going to comment but I had non signal all day. But if you want to simplify the UI, maybe change this to advanced options? And only show them if an advanced check box is checked?

Jaime

Sent from my iPhone

On Jul 29, 2014, at 4:52 PM, chuckd1974 notifications@github.com wrote:

Thank you Brandon! The awesome work you do is greatly appreciated!

Chuck

On Tue, Jul 29, 2014 at 11:22 AM, Brandon Basso notifications@github.com wrote:

Interesting--I guess I can see the argument: cross-track errors are more likely than along-track errors, so you want more sidelap to guarantee good coverage. I'm convinced--lets keep it in there!

— Reply to this email directly or view it on GitHub https://github.com/diydrones/MissionPlanner/issues/542#issuecomment-50540669 .

Charles Devaney Masters Candidate Department of Geography University of Hawaii at Manoa — Reply to this email directly or view it on GitHub.

[iskess]

Thank you Brandon. Maybe we can actually take this a step further and allow the user to enter a prevailing wind velocity. This will affect the overshoot ratios, the calculated max triggering speed, and possibly other calculated values like mission time.

Ultimately I would love to see the ability to compensate for changing terrain by adjusting the width of the flight lines to maintain sidelap. I know I'm really dreaming with that one.

I'm in favor of more options rather than less.

[roque-canales]

I'm also in favor of more options rather than less!

[bbasso]

@iskess Yeah, I think if there were a way to draw a wind vector on the map it could do the following -design a landing approach (plane) -change the grid angle (based on crosswind/withwind radio button for selecting desired direction of travel with respect to the wind) -change overshoots

Recalculating mission times and trigger times is pretty hard, as it depends quite a bit on how the particular aircraft appears to the wind.

+1 for full terrain-based survey. I think this is partially solved with Tridge's new terrain following, which will try to keep a constant AGL altitude rather than always being based off the home altitude.

[iskess]

Calculated max triggering speed would be really easy. Just add the tailwind component of the wind vector to the airspeed. Then the user could see if the required triggering speed exceeds his camera's ability. He can go to the field knowing that if the wind exceeds XX, his will miss shots and needs to realign or cancel the mission. If you add the max triggering speed to the camera database, then it can flag a warning automatically. We already have "Photo every (est):" , we should add a max and min triggering speed so we can see the wind's effects on our triggering speed requirements.

I don't think we ever want to have an automatically designed landing approach. The user needs to be responsible for lining up the final approach manually to assure obstacle avoidance.