All right. I went on to create some reports. First of all I installed the plugin called "Annotation Export CSV, and when i execute the report the same area gets a Volume CYD of 5214 cubic-meters:
After this I went to “REPORT” in the top menu and executed the “Annotation report”, and this is the result:
Conclusion: I have three measurements of the same pile, 3987, 3987 and 5214 qu-meters. Two are reporting the same volume, and the CSV report says 5214 qu-meters and seems to be way off. At least I think so, but… what if this is the correct one, and the two others are off.
Can I trust the two similar measurements to be, at least, reasonably reliable?
I have been using Base plane: Linear fit in all volume measurements, but I really don’t understand the difference between the available Base planes. I compared the results from this session with measurements from the same area created on mapsmadeeasy and “Linear fit” was gave the best similarity. Any words on that?
Thanks for your time, and apologies for my English…
Draw some cross-sections and do a little math to verify. An airplane is for somewhat flat basis. Triangulated is for a little bit of slope and lowest point is if you get up against something that you can’t click around like if you are making a ledge or a wall.
I feel a little uncomfortable having to distinguish between ”somewhat flat basis” and ”little bit of slope”. I’ve tried all three at home on the same area/volume, and the difference is HUGE:
6242 using ”Linear fit”
8480 with ”Lowest point” and
6195 with ”Triangulated”
Linear fit and Triangulated are more or less the same, but ”Lowest point”… I mean: If I pick the wrong base plane the customer will kill me…
Not comfortable at all imho. But I guess I’ll have to get used to it…
Can you explain why the CSV-export gave such figures? Can I trust that the results are at least roughly correct?
R
Lowest Point will always give the largest volume since it creates a flat plane at the level of the lowest point found in the boundary curve you drew. So if the pile is on a slope this will cause a large error. To estimate the slope on which your pile lies, draw 4 profile lines outside of the pile, 2 on either side along its length and 2 on either end along its width. You should notice a significant slope it some of these 4 profile lines. If so you definitely need to use Linear Fit or Triangulated. If the whole pile seems to be on one constant slope, then Linear Fit should work well and give about the same volume as Triangulated. But if the slope is changing along the length or width of the pile, then Triangulated is a better choice; in this case Linear Fit can give a larger or smaller or same result and is not as reliable.
Using the profile lines drawn on the ground just outside the pile will give you a much more solid basis on which to choose one of the 3 base options. The most challenging situation is where you cannot draw a profile line on all 4 sides that is representative of the ground level of the pile next to the line. This happens when a pile has one end up against a cut in a hillside. In this case if you know that the ground beneath the pile is flat, you can use Lowest Point. If not then it may be difficult to get an accurate result.
@SolarBarn: thanks for your clarification. Gee, I wish i could find more to read about this. Do you have a link to share with me?
However. My piles were on a rather flat ground. At least what the Elevation map tells. (The dark green is the shadow) See below:
Triangulated reports 6749 qu-meters
Linear fit reports 7093 qu-meters and
Lowest point reports 8124 qu-meters
No slopes to talk about in the vicinity. Still I had almost 20% larger volume using the “Lowest point” compared with Triangulated. Which makes no sense to me. Ther is no slope around.
About drawing “profiles”: I see your point, but how do I “see” if there is a slope between the profiles? Is there an app available for this? Sorry: I just don’t understand how this should be achieved.
Sorry for beeing such an idiot, but I can’t get this together.
Inside DroneDeploy there is a tool for making the profile. You select the tool, click at the two end points of the line, push enter and the distance between the points is shown along with a profile showing the elevation along the line.
If you can export the .obj file for your site and put it on Google Drive or similar, I can run it thru my Rhino DroneMaps app and will probably be able to give you some details showing why the volumes are different. Just sent me a link, either here or in a PM and I will take a look.
I finally found the tool. Ihave never used the measurement tool before… The bottom under the pile in question moves around 75 centimeters in height along the longest side (75 meters).
Back to my original question: Has anyone tried this “Annotation Export CSV” tool. Nice, but unreliable imho. Or has the error soemthing to do with the rest of my problems?
Double-click on the image to see a larger view. As you can see, the Lowest Point base is below the Triangulated base over the majority of the area, resulting in a volume estimate that is 13% bigger.
This makes sense when you look at the elevation plot which shows the pile to lie on a slope as you noted in your profile plot.
The actual numbers vary depending upon the outline you generate, with the Lowest Point result being particularly sensitive to where the lowest point is located at the lowest-elevation end. That lowest point modulates the whole volume estimate whereas for Linear Fit and Triangulated it is only 1 among many and so has little impact.
Hope this helps you to better visualize what is going on.
Thanks. This is so new to me. I mean: I have done hundred of maps, both Ortophotos as well as DSM, but just recently gone into volume calculations. So this is completely new to me.
You are writing ” As you can see, the Lowest Point base is below the Triangulated base over the majority of the area”. Actually I don’t. The only obvious to me is that the ”Average height” is some 50 centimeters higher than the two others, and .5 times the area (2000 sq meters) gives some 1000 qu-meters which is the difference irl, so to speak. But I can’t see this, so to speak…
Let me think loud: If I have a pile on a slope I have to measure how big this slope is. The ”Lowest point” method will start from… the lowest point and ”build” from there. As in my example it gave a volume 0,5 * 2000 bigger than the two other methods.
But I’m still wondering how I should think when selecting method? How do I know when I’m off?
If I have a pit on an absolutely flat ground I can use either of the three methods?
But let’s say that I have a 75 meter pile. The first 10 meters are -2 meters below, and the rest is ±0. How do I handle this? In my mind the ”Lowest point” can never give a correct useful result without me recalculating the result (Result - 22000 + 210* Width of the pile [Subtracted 4000 plus First 10 meters were two meters down])
I’m sorry If i harass you guys with stupid and somewhat confused questions, but, as I wrote in the beginning, I’m new to volumes.
Yes, on flat ground all 3 give the same result.
On ground with varying slope, triangulated does best because the base more closely follows the slope changes since it is not generating a flat plane like Linear Fit and Lowest Point do. Notice in the side view that the grey triangulated base has some thickness to it since it is bending to follow the varying slope. You need to look at it on a PC with a good monitor and not a mobile device in order to better view this very high aspect ratio picture.
The triangulated method is called this because it fits triangles to the boundary points that you clicked on. So you should use a higher density of points in areas of changing slope in order for the fitted triangles to better model the terraine. The volume is the height of the mesh over these triangles in the base so it is important that the triangles hug the ground well.
@SolarBarn: Thanks for the clarification. I think I’ll have to do some serious field experiments. Good rule of thumb in your postings. Good to have. Also thank’s to @MichaelL for the link. I’ve been looking for this, but haven’t found it.
