Welcome to PhantomPilots.com

Sign up for a weekly email of the latest drone news & information

  1. Airy1991

    Joined:
    Aug 7, 2018
    Messages:
    1
    Likes Received:
    0
    Hello there,

    I am writing today because I have so many questions about drone accessories, photogrammetry and orthorectification.
    First of all, I have never flight a drone, but I have taken GIS and some basic clases of remote sensing, where they taught me it is necessary to have al least 4 ground control points in order to Orthorectificate an image. Now, my problem is that I am a marine biology student designing a study in the middle of nowhere in the sea, where there are not distinguishable ground control points. I will be trying to do photogrammetry to measure the lengths of manatees, but also counting how many manatees I see during my flights.

    I will be flying the Phantom 4 (according to my advisor) in a straight line and trying to take photos that overlaps. The only GPS position I will have will be the center of the picture (location where the camera was when I took the picture). Also, I know the drone have a barometer attached but is not precise (al least that is what my advisor said). I mention this last thread because if I can not orhtorectified the pictures, in order to measure lengths of the animal I will need to know the altitude, and my advisor said it is not possible with the barometer because they were not able to measure the true lengths before. I was thinking that perhaps they were not aware of the distortion of the camera and that is why I want to try to orthorectified the images

    My questions are:
    1. ¿Is there any way I can orthorectified image if the only location I know is the center point of the image (Where the camera was at the moment of taking a picture)?

    2. I will be doing line transects, which means I will be flying the Phantom 4 in a straight line and taking pictures that will be overlaped, would this help for Orhtorectification? How? and which softwares should I use for it?

    Finally,
    3. In order to do photogrammetry, is there any other accessory that you would recommend me to add to the drone besides a barometric pressure sensor for measuring altitude? I was reading about LiDAR for drones, but I do not know if that would work in water.

    Thank you for your time,

    Airy
     

    Attached Files:

  2. WV. Rootman

    Joined:
    May 1, 2017
    Messages:
    1,508
    Likes Received:
    839
    Location:
    N.central wv. U.S.
    I am a nobody. I have read of companies that can do aerial measurements. Check past posts for more information. They were asking the same question you are.
    My suggestion is throw a known target in the water before you fly the drone. No matter what elevation you are flying, as long as the target is in the picture, you can get your measurement.
     
  3. Meta4

    Meta4 Moderator
    Staff Member

    Joined:
    Jan 8, 2014
    Messages:
    12,675
    Likes Received:
    6,929
    This sounds like a very interesting project but there are some serious obstacles that will make it very difficult to carry out.
    To produce orthophotos, the software that is commonly used identifies hundreds of common points in each pair of overlapping photos.
    On land, this is mostly fairly easy for the software to manage but some areas have featureless or confusing ground patterns like grassland or forest canopy and that can cause problems.
    Where the ground pattern is water surface, there will be almost no fixed points that are identifiable between photo pairs and the photogrammetry software won't be able to do much at all.

    I'm not sure you could use photogrammetry software to produce orthophotos of the sea surface and I'd be wanting to clarify that with your supervisor before getting too far out on that limb.
    You could easily use a mapping app to fly precise transects and capture images along the transects.

    On land, you can produce reasonable good orthophotos just using the stored image data and the triangulation calculations the software will perform on thousands or millions of identified common points from the images.
    You only need to use ground control points if you want to get accuracy, particularly for heights and be able to match the local height datum.
    Without GCP's the heights will just be relative to the starting point.

    That is the usual method.
    For mapping projects, there are specialised mapping apps like DroneDeploy that take care of the planning, flying and image acquisition that would be otherwise impossible.
    Here is a screenshot from DD when flying a 32 acre site and collecting images with a precise 78% front and side overlap.
    [​IMG]
    Here's a plot of the centre points of the captured images:
    [​IMG]
    Agisoft Photoscan is the most commonly used software for creating orthophotos from the images collected.
    Or the companies that produce mapping apps will also do the work for you at a cost.

    The drone will use a barometric sensor for flying altitude and DD will maintain altitude very well using this.
    But oddly DJI record the GPS altitude in the exif information with each image and GPS is inaccurate for altitude and the altitude recorded may vary considerably during a flight even when the drone stays at a constant height.
    Here's an example of what I mean:
    [​IMG]
     
Loading...
Similar Threads - Manatees research project Forum Date
Distance, distance,distance...Doing research and design General Discussion Jul 5, 2018
Getting better All In Good Time. Thanks for your research help. Photos and Video Jun 14, 2018
New to drones - Researching where to purchase DJI Mavic Air - DJI or dealer? Introductions May 1, 2018
More plane vs quad research..... General Discussion Dec 14, 2017
New pilot researching pilot error General Discussion Oct 11, 2017