jdow

Members
  • Content Count

    9
  • Joined

  • Last visited

Community Reputation

0 Neutral

About jdow

  • Rank
    Newbie

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Kevin: Yes;, you are correct. There's a "c" left out of the spelling. Please try the following link: http://www.immersivegeomaticssolutions.com I tried to add some examples here, but the file sizes were too large. Thanks for bringing this to my attention.
  2. Kevin: For examples of curved flight lines, take a look at my website at (http://www.immersivegeomatissolutions.com). Look at "How we work / FODAR / Curved Flight Lines".
  3. Kevin: I agree with your experience on all counts. Also, I have found that the addition of oblique (off nadir) images definitely helps; and adding overview obliques helps more. The impetus for using curved flight lines has to do with the SfM processing propensity to accumulate the lens calibration error over image blocks when traditional (linear/parallel) flight lines are used to collect the imagery (several researchers have verified this). It's just too easy to use curved flight lines to eliminate the accumulation of lens calibration error. If linear/parallel flight lines are being used; then why not curved (non-linear/non-parallel) flight lines? There are other benefits of using curved flight lines, especially when fixed-wing drones or piloted aircraft are deployed - like minimizing turnout loops (time, fuel/battery, costs). For quadcopters, there's no need to pause, rotate, and proceed down a short leg only to pause, rotate, and proceed down the next parallel flight line - just keep on flying around the curved loop.
  4. My experience runs parallel to that of R Martin on the subject. Give it a rigorous trial before you spend the money. Also, be aware of the well known systematic Structure-from-Motion doming/dishing (elevation) error. Just Google "Structure-from-Motion doming/dishing error" and take a good look. That's why I am skeptical of the solution you describe. Right now, I'm trying to get good acceptable results (without spending the big bucks on "direct geo-referencing" equipment) using consumer grade GPS and cameras while mitigating the SfM doming/dishing (elevation) error by using gently curving, convergent, non-traditional (non-linear/non-parallel) drone flight paths along with some overview oblique photos.
  5. Kermange: If you are seeing points "underground" or even "above ground", You may have encountered the well known systematic Structure-from-Motion doming/dishing (elevation) error that most (PhotoScan and Pix4D) SfM packages are capable of producing when the images are collected using traditional (linear/parallel) flight lines and the camera lens calibration is not spot on - search the research literature from 2014 (James) forward. Collecting some overview oblique photos along with the normal image collection helps as well; and good ground control is a must have. I'm switching to gently curved, convergent, non-traditional (non-linear/non-parallel) flight lines to help mitigate the SfM doming/dishing (elevation) error - flight lines being flown using fixed-wing drones and piloted aircraft.
  6. AgiSoft (PhotoScan), Pix4D, and Skyline (TerraExplorer) are packages that I am basically familiar with - from testing and comparing. All three are good. However, be aware of the Structure-from-Motion systematic doming/dishing (elevation) error that can occur when using traditional (linear/parallel) flight lines. Add some overview oblique photos to help out the process. I have found that even some of the service bureaus who process the images don't know about this type of elevation error - which occurs frequently. Be careful. I'm trying to switch to gently curved, convergent, non-traditional (non-linear/non-parallel) flight lines with oblique images to try and mitigate the SfM doming error.
  7. It's not just California! It's Mississippi, Florida, North Carolina, etc. It seems to be directed at protecting land surveyors who just haven't gotten into the technology yet; but it's working. It's best to work through a licensed land surveyor who can ground truth and stamp the positioning (especially elevation) results. However, be aware of the well known Structure-from-Motion systematic doming/dishing (elevation) error; or it will bite you as well as give the land surveyors a silver bullet.
  8. If you are doing work for rock quarries, land fills, etc., what has been your experience with regards to the systematic Structure-from-Motion doming/dishing (elevation) error? I have switched to gently curved, convergent, non-traditional (non-linear/non-parallel) flight lines to mitigate this well known error. Oblique photos help as well.
  9. Has anyone on this forum tried using curved, convergent, non-traditional (non-linear/non-parallel) flight lines to gain a more diversity of view perspectives? Additionally, these non-traditional flight lines (along with oblique photos) tends to mitigate the systematic Structure-from-Motion doming/dishing (elevation) error. Visit my website (http://www.immersivegeomaticssolutions.com) for examples. Access "What we do / FODAR / Curved Flight Lines".