Support Website Contact Support Blog

Ground adjustment problem

Hello dear Pix4d community,

Currently, I am working on a project that requires high accuracy on elevation profile values (Z-axis values). The problem is that once the map is constructed by the software, I generally observe misleading information across the height values. Although, in reality, my measurements on two different position’s height values are approximately the same; I find out in the software that they are much different. i.e. (40-50 meters offset).
At first, I thought that this situation might be reasoned from the occurrence possibility of some of the faulty or deceitful aerial images in my data set, which I have gathered from different drone flights, but then I find out that it does not only happen on the same place, but also appears on different places or even different maps.
After analyzing further, I determined that the map is lifted (or raised) 10 degrees to 20 degrees from one side to another, or in other words, rather than observing a flat line (height values should not be changing much once I move the cursor on an imaginary line), I observe increasing or decreasing values, which are approximately advanced by tangent 10 to 20 degrees.
I’ve done a search on google, and gathered some information about the GCPs (Ground Control Points), but I could not manage to implement that. I think that my problem will be solved once I correctly place those gcps. Therefore, Can any one help me by giving an example to place those points appropriate, and construct a reliable map?


Hello Serkan,

What you describe can be solved by adding GCPs as you mention. 

You can find everything about GCPs within Pix4D here.

If you want to know about how to distribute them, how many you need, etc, please find more information here.

In case that for whatever reason, you cannot measure GCPs and you want to avoid the problem of your project being lifted, you could also use RTK/PPK geotags in your projects. That means that very accurate geotags will be used and therefore they will have much more weight in the computation than your standard geotags and the project will not lean or undulate. More information about this topic can be found here.