Hello Fellow Yuneec Pilot!
Join our free Yuneec community and remove this annoying banner!
Sign up

Manual camera settings for a survey/corridor mission

Joined
Jul 11, 2018
Messages
211
Reaction score
77
Location
Poland
Hi all,

Have you ever flown survey mission/s with manual camera settings? I have tried and have a feeling that my efforts failed. Images meta show that despite my efforts i.e. setting manual shutter speed and ISO etc. in DataPilot all pictures were taken in Auto mode anyway. Yes, I clicked "No" when prompted whether I wanted to perform the mission with recommended DataPilot mission settings.

Any ideas on how to make Manual settings work? What is a workflow?

Pawel
 
Joined
Sep 1, 2016
Messages
180
Reaction score
74
Age
52
Hi all,

Have you ever flown survey mission/s with manual camera settings? I have tried and have a feeling that my efforts failed. Images meta show that despite my efforts i.e. setting manual shutter speed and ISO etc. in DataPilot all pictures were taken in Auto mode anyway. Yes, I clicked "No" when prompted whether I wanted to perform the mission with recommended DataPilot mission settings.

Any ideas on how to make Manual settings work? What is a workflow?

Pawel
Good question, but I am afraid that there is no shutter speed or ISO value priorization possible. Both in combination could be used to trade noise for blurriness... i.e. on low light occasions.
I asked for it a long time ago, but it did not come.
I use WB Cloudy, EV+/-0, Ultra JPG for the E90 and set "No changes" in the Mission Planer's camera settings. This way at least AWB is disabled.
 
Joined
Jul 11, 2018
Messages
211
Reaction score
77
Location
Poland
So I did a few flights (surveys) earlier today.
I did two sets of flights of the same polygons: one - at Manual Exposure and the other at Auto Exposure.
When Manual exposure was in use the camera displayed, while in flight, that the e90 was keeping pre-set values (manual: shutter - 1/2000s and ISO 100). Having checked exif data I cannot see it was: most of the frames were exposed at 1/961 but ISO was bouncing somewhere between 112 and 153 or so. Yet a few pictures were taken at 1/1923 and ISO 183 or 165.

Can anybody say what is a workfow which would secure that Manual Exposure settings work for pattern missions?

I sadly add that I was flying Mavic 2 Pro and Phantom 4 Pro v2 the other day and cameras settings there worked great. A pilot has a full control over exposure settings and has plenty to chose from: Auto, Aperture, Shutter and Manual and furthermore can also set ISO at Auto or Manual using either of the exposure modes. Wish Yuneec neared that.
 
Joined
Sep 1, 2016
Messages
180
Reaction score
74
Age
52
When Manual exposure was in use the camera displayed, while in flight, that the e90 was keeping pre-set values (manual: shutter - 1/2000s and ISO 100).
What I additionally noticed as major bug in manual exposure mode, that when you interrupt the mission i.e. by stick input, the exposure time gets set to the longest possible duration of 4 seconds. What you then first see is a "white out" ST16S display because daytime light conditions causing a fully overexposed image. So you have to correct the exposure time setting mannually in order to regain a normal exposed view...
 
  • Like
Reactions: Pawel_(PL)
Joined
Jul 11, 2018
Messages
211
Reaction score
77
Location
Poland
Having asked Polish Yuneec team (I thank them) got instructions on how to sucessfully set ME in Pattern (survey/corridor) missions.

In order to effectively set ME we need to tick Photo from drop down menu under Mission Start/Camera (pls. see attachement received from Pl team). Survey is set by default which needs to be changed for Photo.

In case the system asked whether to enable default camera settings chose No. Mine does not ask...

Flew a few missions and it really worked - images were written with manual settings.

Hope to help those interested.
 

Attachments

  • 20210811_135501.jpg
    20210811_135501.jpg
    123.2 KB · Views: 7
Joined
Sep 1, 2016
Messages
180
Reaction score
74
Age
52
Having asked Polish Yuneec team (I thank them) got instructions on how to sucessfully set ME in Pattern (survey/corridor) missions.

In order to effectively set ME we need to tick Photo from drop down menu under Mission Start/Camera (pls. see attachement received from Pl team). Survey is set by default which needs to be changed for Photo.

In case the system asked whether to enable default camera settings chose No. Mine does not ask...

Flew a few missions and it really worked - images were written with manual settings.

Hope to help those interested.
Thank you for this promising info! I might give it a try tommorrow if weather permits some useful AG surveys (with photo mode and ME settings)....
 
Joined
Jul 11, 2018
Messages
211
Reaction score
77
Location
Poland
Thank you for this promising info! I might give it a try tommorrow if weather permits some useful AG surveys (with photo mode and ME settings)....
Welcome and good luck.

Please share your thoughts afterwards.

What AG surveys do you do if i may ask? What kind of data and analysis do you do?
 
Joined
Sep 1, 2016
Messages
180
Reaction score
74
Age
52
Welcome and good luck.

Please share your thoughts afterwards.

What AG surveys do you do if i may ask? What kind of data and analysis do you do?

Unfortunately "Photo Mode" and manual exposure settings still cause a "white out" screen (4 seconds over exposed image/screen) when interrupting the DataPilot mission i.e. with a stick input:
Screenshot_2021-08-18-17-59-51-edit.jpg

Regarding AG surveys: I have an NDVI-camera (12MP modified RGN) on a Phantom-3-Pro and testwise an NRG-sensor (quiN) for the H520. Additionally I use the E90 RGB imagery i.e. for comparison and put all the stuff into DroneDeploy AG-Plan and sometimes Agremo, tested Agisoft Metashape (manual filtering) and earlier on also Pix4D with different modules for about three years, but they often had problems stitching corn fields of any type I fed them with. Most reliable stitcher for my uses is DroneDeploy, that is why I stick to them.
Actually crop analysis (via drone aerials) is much more than a science of its own. Especially the core essential "sun direction and the many kinds of different reflection effects" (i.e. halo/glint) make it a hard task, even on cloudy days...
I would not go so far as it does not work, but it is one (further) sector of former promising wide spread drone application that was quite overrated over the last years. The results that everyone desires can only be reached with sensors > 50,000 EUR... Or how SlantRange recently promoted in its newsletter: "more efficient research & breeding", and that is for the time beeing also my conclusion: AG drone application might be useful for researches and breeders, when operating on small fields with high end sensors and good counter samples from ground staff...
 
Joined
Jul 11, 2018
Messages
211
Reaction score
77
Location
Poland
Hi,

Thanks for feedback.

As for 4 sec. issue it has not been a problem for me so far for survey missins are performed in automated mode anyway. I admit that exposure turns to 4 sec but only in landing when it is irrelevant. Otherwise my priority has been to "force" the system to take pictures in ME mode and maintain this during the lenght of a mission and write such images on a card. The described method seems to provide for that. I have never had a need for i.g. stick intervention while in survey mode. But have a feeling that there should be enough time to reverse4 sec effect and go back to desired ME settings. Still while in flight.

As for the AG surveys and a need for them. My feeling is that a problem with scaling them results from among others: 1. lack of education among farmers, 2. farmers' need can be fully met by satellites what is easier and saves them money spend on us travelling from site to site.

Agree thst R&D is where we can have business and here where we outperform satellites.

Having said so I do feel that hardware does not need to break 50 k €/$. High quality multispectral is around 5-10k and this is sufficient to meet R&D needs, at least today. RGB shoul not exceed that.

Br,
Pawel
 
Joined
Jul 11, 2018
Messages
211
Reaction score
77
Location
Poland
Maize is really tricky to produce good maps. It's very homogenous when looked from atop. Therefore it is advisable to increase overlap to 80 or even 85% and fly higher if possible. And keep fingers crossed it works.

Multispectral - We use radiometric calibrated Parrot Sequoia and Micasense RedEdge and they produce good enough results to be used by R&Ds of various companies as well as farmers (V.I. based zone maps for variable rate applications, biomass, vigor, overall condition, stress detection, etc.)
 

New Threads

Members online

Forum statistics

Threads
19,921
Messages
230,284
Members
24,106
Latest member
Griggsrj