I am looking to do a project for a local excavator at home sites that are to be leveled and plotted. These homes are on a pretty steep hill, so there is a significant amount of dirt that needs to be pulled down to level the plot, and then also to be hauled away. We want to find the amount of dirt that exists on the property above the leveling datum and how much dirt is needed to fill in the area below the leveling datum. I have a few questions.
First off, will the Autel Evo 2 Pro (6k images, 1 inch sensor) work with providing the images and data needed for a successful map? (do i NEED an RTK drone?)
Secondly, would i want Ground control points at the datum (the point at which the map should be level)? I am not really sure how GCP’s work or how i would use them, but i do know it will increase the level of accuracy.
Third, will ODM calculate the volume of dirt in an area above the datum, and also the negative volume of empty space needed to be filled below the datum?
Thanks in advance for your tips and knowledge! I am new to the mapping and drone data system, and am hoping to provide a great experience for the company.
That doc entry looks like it will help, but in all honesty I am trying to understand everything at once. Is there a tutorial video on collecting data and combining it into a map? Like a start to finish operation?
I really appreciate the help and can’t wait to get to work.
(I can probably look this up, but how would you explain PointCloud to an idiot because that’s me haha)
Ooof… Probably, but not something we’ve made. Our docs should get you most of what you need, or at least point you the right way. A start-to-finish guide with pointers is on my radar to draw up…
Point Clouds are pretty simple stuff in reality, I think! In effect, nothing more than what can be represented in a spreadsheet, really, having X/Y/Z and R/G/B (and other values like classification) for each point. That gives you the 3D location and attributes (color) for every point. We use PoTree to optimize this data in a special format and display it interactively/quickly in the browser, and to bring along some awesome tools like measurements, etc.
It gets reconstructed via photogrammetry (basically comparing multiple images and using image differences to get depth/height measurements) and is the basis for the 3D model we generate.