Assistance with data sets for Mosaics

Good Afternoon!

I am following up to my previous post looking for the viability of the parameters in my data set for mapping.

I appreciate the assistance provided by Saijin_Naib, Gordon, and bernarde. I have made adjustments to the ISO and Aperature settings on the camera and changed the gimbal angle to 85 degrees.

Instead of an orthophoto, I am looking for more of a stitched image of my data set, and the process for doing such an image. I believe this is called a mosiac? Below I will provide links to an example image as well as my most recent data set.

Example photo:

Photo set: Dropbox - 2August2021 - Simplify your life

1 Like

An orthophoto is an orthorectified image mosiac.

What are you looking for exactly, and why is the orthophoto product not what you need?

QGIS can mosaic, but you might have issues with seams.

1 Like

That’s a great question, I just noticed that the mosaic example references the orthophoto and that you may be correct as to the ortho photo being what I am looking for.

I’m looking for a method to stitch the photos together and was wondering if there was another method you may commonly use.

Or, if that is the best way to get the photos stitched together, is the current data cleaner?

1 Like

In very rare cases I’ll generate a naive 2D stitch using something like Microsoft Image Composite Editor, but that’s more just for validation that the data should be able to process.

In general, for most people’s applications, I think the toolset provided via OpenDroneMap for orthophoto generation should be the best course.

1 Like

Thank you Saijin_Naib! Then my other question is, do you see any room for improvement in the photo process?

Sure! We can always do things better, I think.

What exact steps is more Piero’s wheelhouse as a lot of it is beyond me.

1 Like

My question may be more related to the image output. I’m looking to have the image output at a high resolution and for the areas above the house and above sections of the trees be more more filled in with the photo data with less white space.

Here is an example image (although it is not currently in it’s highest resolution form)
I also recognize that this image has more distinct markers outside of trees, one house, and a water feature.

And here was one of the outputs from a previous data set - report.pdf (8.7 MB)

1 Like

Being more filled in is (almost) entirely up to the flight plan and having sufficient overlap/sidelap.

Ah, do not think of water features as a positive thing in photogrammetry. We as people can recognize them from the air readily, sure, but stitching them is borderline impossible, especially if they’re spectral at all.

You did a strip survey and got honestly pretty awesome coverage over your strip. If you want to “fill in” more, you might need to combine your strip with a typical grid survey along the strip as well, to capture more angles of each feature to help in-fill more during reconstruction. Also, shooting slightly off-nadir would likely help a bit.

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.