Forum

Author Topic: Combining terrestrial (ground based) photos with aerial (uav) images  (Read 34021 times)

Yoann Courtois

  • Sr. Member
  • ****
  • Posts: 316
  • Engineer in Geodesy, Cartography and Surveying
    • View Profile
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #15 on: May 16, 2017, 03:54:53 PM »
As you said, P4P camera is a lot better than previous Phantom cameras.
I of course red with interest your link and learnt a lot about why this camera is almost consider as a frame sensor, even if inner parameters (focal length...) look similar to fish eye lens (such as GoPro).
But you have also to consider that number of pixels is not everything :
- Most of people turn the nose up at fish eye lens for photogrammetry, I was able to handle measurement with Z-accuracy of 0.8mm using GoPro (400€ device)
- You seem to be puzzled about P4P sensor, I was able to handle measurement with Z-accuracy of 0.4mm using P4P (1500€ device).
=>  Both of those surveys where compare to a Leica TDRA6000-station measurement
Those devices are not able to make centimeter accurate measurements from pictures taken at 100m heigh, but when you use it well, you can be pretty surprise about it ;)
--
Yoann COURTOIS
R&D Engineer in photogrammetric process and mobile application
Lyon, FRANCE
--

Packy_GIS

  • Full Member
  • ***
  • Posts: 107
    • View Profile
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #16 on: May 16, 2017, 10:53:31 PM »
I completely agree - I made my own tests with GoPro 3 and if you handle the handle the data well (and use regularely spaced GCPs) you can get very usable results.
I even use Kite in combination with GoPro to create 3D models of archeological sites in Sudan.

Jan

Yoann Courtois

  • Sr. Member
  • ****
  • Posts: 316
  • Engineer in Geodesy, Cartography and Surveying
    • View Profile
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #17 on: May 17, 2017, 10:12:46 AM »
Kite ???
Is this local UAVs ?  ;D

By the way GoPro 4 has really surpass the GoPro 3 quality. I'm really raring to work with the Hero 5 model.
« Last Edit: May 17, 2017, 10:24:19 AM by Yoann Courtois »
--
Yoann COURTOIS
R&D Engineer in photogrammetric process and mobile application
Lyon, FRANCE
--

Shovelhead

  • Jr. Member
  • **
  • Posts: 75
    • View Profile
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #18 on: May 17, 2017, 09:55:45 PM »
Hi everyone !
I'm actually doing my master thesis in photogrammetric methods for topographic surveys.

Some tips I could give you:
- Integrated GNSS data are not mandatory: Indeed they are never better than a couple of meter accurate. So you will need well-known GCPs to accuratly georeference your model. I usually untick pictures georeferencement when handling survey with our Phantom 4 Pro (Really nice camera by the way).
- Ground pictures will permit you to fill your holes under trees: I'm actually using GoPro camera at the top of a pole. Even everybody are scared about fisheye lens for photogrammetry, I can tell you it's possible ! I was actually able to handle super high accurate measurement using GoPro (I mean sub-millimetric accuracy)
- The secret is recovering: The more your pictures recovered each other, the strongest will be your model. And don't be scared about number of pictures. I've actually noticed that a stronger bloc with a huge amount of pictures can be calculated faster than a smaller number of pictures weakly linked. (Here is where GoPro camera makes it stronger because of the laaarge field of view)
- Coded targets give you the possibility to save a lot of time in processing: and will give you more confidence on reconstruction. I postpone you to my topic talking about coded target size before print them.

Hope it helps !
Regards

I really appreciate your reply, any other tips, tricks or examples you'd like to share. I'm especially interested in coded targets.

Packy_GIS

  • Full Member
  • ***
  • Posts: 107
    • View Profile
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #19 on: May 17, 2017, 11:15:03 PM »
Kite ???
Is this local UAVs ?  ;D

By the way GoPro 4 has really surpass the GoPro 3 quality. I'm really raring to work with the Hero 5 model.

Well - I use KITE (KAP) when I need to take aerial imagery and the UAV is not available or the usage is prohibited - I can bring kite to Sudan without any problem, but if you bring an UAV - then be prepared for a long lasting administrative fight with the customs. On the other hand the UAVs are restricted by law in my country - and Kites are not.

HERE you may find my paper dealing with the Kite survey a little.

Jan

Yoann Courtois

  • Sr. Member
  • ****
  • Posts: 316
  • Engineer in Geodesy, Cartography and Surveying
    • View Profile
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #20 on: May 18, 2017, 09:26:35 AM »
Hi Shovelhead !

Here is as topic where we were talking about coded-target size. As I didn't get any precise answer, I've handled test for my personnal use which I've shared the results. No problem to answer you if you have other questions !

Hi everyone !

After managing some tests with different cameras (UAV DJI Phantom 4, ActionCam GoPro Hero 4, ...),
I was able to define the minimal size a coded target should have in order to be detected !

Tests parameters :
I have handled my tests with 12 bits target, because they present, according to the user manual, the most ability to be detected.

Result :
In order to be detected, the center point of a PhotoScan 12 bits target have to cover a minimum of 9-10 pixels on the image

Application :
For instance, for a photogrammetric survey with a defined GSD of 1cm, the center point of the target should be around 10cm minimum (10 x GSD). As the global diameter of the target is 3.5 times the diameter of the center point, the whole target should have a minimum size of 35cm. (35 x GSD).

To be printed :
As the center point diameter should be egal or greater than 10 x GSD, the Center point radius should be set at 5 x GSD or more.

For non-normal view :
I didn't handled precise measurements but I can say the angle of view is not the main limitation in automatic detection. For UAV survey with nadiral images for example, the limit of detection will be the effective GSD on the target when the camera will fly away from the target position, and not the angle a view.

Hope it can help  ;)
--
Yoann COURTOIS
R&D Engineer in photogrammetric process and mobile application
Lyon, FRANCE
--

Yoann Courtois

  • Sr. Member
  • ****
  • Posts: 316
  • Engineer in Geodesy, Cartography and Surveying
    • View Profile
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #21 on: May 18, 2017, 09:29:52 AM »
Hi Jan !

Your resourcefulness is wonderful ! Never though that kind of process were possible !

And thank you very much for sharing your papers. I'll proudly use them in my master thesis bibliography !

Regards
--
Yoann COURTOIS
R&D Engineer in photogrammetric process and mobile application
Lyon, FRANCE
--

Packy_GIS

  • Full Member
  • ***
  • Posts: 107
    • View Profile
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #22 on: May 18, 2017, 01:41:17 PM »
Hi Yoann!

Here is a video where is one of my students kiting. Neverthless to say, that about 10% of images are well used for 3D modelling.

Even if we are a small university, we are trying to do good and interesting research - and the 3D modelling is just a small part of my research radius :) I deal with old maps, archival aerial photographs, web mapping apps, cloud GIS services, LIDAR collection and processing, remote sensing .... ufff and so much more ;-D

You can be my student in Erasmus programm - I have lessons in English as well  ;D ;D

Jan

Yoann Courtois

  • Sr. Member
  • ****
  • Posts: 316
  • Engineer in Geodesy, Cartography and Surveying
    • View Profile
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #23 on: May 18, 2017, 02:38:54 PM »
Thanks for this ! I can realize how it looks !

Ahah thank you for such offer ! I've already gone to Helsinki University for Erasmus 2 years ago.
I'll have finish my studies by the autumn ! But hopefully I will continue developments around photogrammetry using PhotoScan, so will still be active here !
--
Yoann COURTOIS
R&D Engineer in photogrammetric process and mobile application
Lyon, FRANCE
--

Jeremiah_ROWE

  • Guest
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #24 on: May 18, 2017, 06:48:28 PM »
My understanding is that image sets from different cameras should be processed separately and then aligned/merged by the GCPs/markers afterwards. This holds true with UAV data from different flight altitudes, so should pole raised grounds shots.

For a prime lens you want minimal distortion and a large aperture. 35mm and 50mm equivlant lenses seem best suited for this, but don't hold me to it!  :)

I just added the terrestrial photos and did the workflow over again (didn't have to pick my gcps again). I plan to purchase a prime lens asap for better results. I was thinking 35mm, is there anything else I should look for in a prime lens?
Also am I supposed to be processing both sets of images separately due to the different camera calibrations or does photoscan automatically take care of that?
« Last Edit: May 18, 2017, 07:01:59 PM by Jeremiah_ROWE »

Yoann Courtois

  • Sr. Member
  • ****
  • Posts: 316
  • Engineer in Geodesy, Cartography and Surveying
    • View Profile
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #25 on: May 18, 2017, 08:03:51 PM »
Dear Jeremiah and Shovelhead,

To my mind you're wrong. Indeed PhotoScan recognizes the camera used for each picture you import in a single chunk, so it creates 2 different cameras if you import both terrestrial and aerial pictures at the same time. (Even better to enter precalibration parameters if available).
And if you gather all the conditions needed (Overlapping / Computer calculation capacity) to align everything at the same time, DO IT !
Indeed, if you're able to do that, you will be able to balance everything at the same time (calibrations, tie points projections, GCPs coordinates) and your model will become stronger !

If something make this process impossible (I've handled a measurement where my terrestrial set of picture hasn't enough overlap with the aerial survey), then you will have to process each survey separately and then align your chunks together using known GCPs or distinctive point you're able to measure in your main model and then export them into your other one.
But when you align chunks together, only a 7-parameters transformation is available (3 translations, 3 rotations, 1 scale factor) so you won't be able to fix any residual distortions.


About lens, I would say it depend on the kind of survey you handle. I usually use micro-sensors (Phantom 4 Pro UAV, GoPro Hero4) with a reasonable flight height (respectively 30-35m and 3-4m) and I'm able to produce centimeter accuracy models.
« Last Edit: May 29, 2017, 03:38:02 PM by Yoann Courtois »
--
Yoann COURTOIS
R&D Engineer in photogrammetric process and mobile application
Lyon, FRANCE
--

Shovelhead

  • Jr. Member
  • **
  • Posts: 75
    • View Profile
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #26 on: May 19, 2017, 05:48:30 PM »
Hi Shovelhead !

Here is as topic where we were talking about coded-target size. As I didn't get any precise answer, I've handled test for my personnal use which I've shared the results. No problem to answer you if you have other questions !

Hi everyone !

After managing some tests with different cameras (UAV DJI Phantom 4, ActionCam GoPro Hero 4, ...),
I was able to define the minimal size a coded target should have in order to be detected !

Tests parameters :
I have handled my tests with 12 bits target, because they present, according to the user manual, the most ability to be detected.

Result :
In order to be detected, the center point of a PhotoScan 12 bits target have to cover a minimum of 9-10 pixels on the image

Application :
For instance, for a photogrammetric survey with a defined GSD of 1cm, the center point of the target should be around 10cm minimum (10 x GSD). As the global diameter of the target is 3.5 times the diameter of the center point, the whole target should have a minimum size of 35cm. (35 x GSD).

To be printed :
As the center point diameter should be egal or greater than 10 x GSD, the Center point radius should be set at 5 x GSD or more.

For non-normal view :
I didn't handled precise measurements but I can say the angle of view is not the main limitation in automatic detection. For UAV survey with nadiral images for example, the limit of detection will be the effective GSD on the target when the camera will fly away from the target position, and not the angle a view.

Hope it can help  ;)

I recall reading somewhere that coded targets should not be used for GPSed GCPs. Was that information false? Also how are you making a rigid target that won't get destroyed right away as I'm sure paper won't last long at all. I was thinking printing one off and pasting it on an equal sized piece of plywood and sealing it with clear coat?
« Last Edit: May 19, 2017, 05:51:23 PM by Shovelhead »

Shovelhead

  • Jr. Member
  • **
  • Posts: 75
    • View Profile
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #27 on: May 19, 2017, 05:57:52 PM »
My understanding is that image sets from different cameras should be processed separately and then aligned/merged by the GCPs/markers afterwards. This holds true with UAV data from different flight altitudes, so should pole raised grounds shots.


I just added the terrestrial photos and did the workflow over again (didn't have to pick my gcps again). I plan to purchase a prime lens asap for better results. I was thinking 35mm, is there anything else I should look for in a prime lens?
Also am I supposed to be processing both sets of images separately due to the different camera calibrations or does photoscan automatically take care of that?

Why do you think you need to process image sets separately?

Packy_GIS

  • Full Member
  • ***
  • Posts: 107
    • View Profile
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #28 on: May 19, 2017, 09:45:31 PM »
There is no need to process them separately - just load them all into one project and let PS do the work.
In 99% I get well usable results.
Jan

cadm8

  • Jr. Member
  • **
  • Posts: 74
    • View Profile
Re: Combining terrestrial (ground based) photos with aerial (uav) images
« Reply #29 on: May 31, 2017, 10:57:05 AM »
A question regarding terrestrial photos... We have a set of aerial photos that align easily and some ground photos (same place) from a number of different locations that do not align by any setting available (even between them). Is there a way to insert manual tie points for the ground photos and then align them with the aerial ones?