Mavic 2 Pro: Flying at 300 meters but DTM produced negative contour line.

PIX4Dmapper PIX4Dmapper Questions/Troubleshooting
1

Hi there, 

Hoping to get a timely reply on the following issues/workaround:

  1. DTM (after classification) generated by pix4d is giving me -ve value for contour line. Mavic pro 2 flew at the height of 300m and register altitude of on average 350m. but when DTM was produced, the lowest point in the map is about -50 meter. 

  2. I’ve been using a preview version of Pix4d which has helped to fix the problem of coordinates. is there a bug somewhere?

  3. I did not collect GCP, and web map services could not produce an accurate result. as such, i am hoping to seek help other then what has been previously suggested. 

TQ

2

Hi Chin Chye Lim,

  1. That’s strange. At the moment, I cannot tell why this happens.
    There is some known issues with the Mavic 2 Pro:
    Pix4D Mavic 2 Processing Issues
  2. It should have been solved in the preview version. However some bugs can always occur.

If you want us to further investigate, please upload your quality report, *.log file, *.p4d file and if possible all images from your project here:
OneDrive

Best,

3

Dear Marco, 

I have uploaded all the files into the drive.

The whole point is to be able to produce a correct DTM model/terrain profile/mosaic map for the whole area as uploaded in OneDrive. I have to use an open source method to produce the terrain model instead of relaying on results produced in pix4d for now. 

Apart from having a problem with -ve contour/DTM, merging projects have been proven very difficult without GCPs (although i have many MTPs laying on overlapped areas but not calibrated when merged). I have repeated the merging efforts too many times and each time i have to delete and redo it again and again but to no avail. it’s very stressful and lost not knowing how to merge properly.

There are questions such as:

Before merging (on individual sub-project level):

  1. Should individual project’s camera setting remain original (not optimised) , & internal camera set to ALL (NOT ALL PRIOR), as long as they pass quality check. What if i change one of the sub-project’s internal camera to ALL PRIOR & also its loaded optimized camera setting (due to 90% relative difference in camera optimization). Would all these settings affect merged project’s accuracy? should them be corrected on the individual project level, or only correct them while merging the project? 

While merging:

  1. Should we change ALL sub-projects’ internal camera setting to ALL PRIOR for ALL, or only for those affected with misalignment of dots? 

2.Should we also change camara optimisation for all back it ORIGINAL, or should be use 

When it comes to merging, initially i came across a problem with misplaced red and blue dots on one of the sub-projects (total 5 to be merged), but it was resolved by changing all sub-projects’ internal cameras to “All prior”. Subsequently, i also changed the camera model back to original (instead of the optimized one) for all the sub-projects in the merging window. BUT this resulted in 4/5 sub-projects’ point cloud appeared in RED, forming RED sea, and failed. 

I seriously need your team to look into those problems as i am hitting the wall now not knowing what to do next to improve DTM, & Merging. 

TQVM and please help. 

 

4

Hi Marco, 

Wondering if you have found out why DTMs were produced with -ve values given those dataset? 

TQVM

 

5

Hi there,

I have been waiting for your reply as to how to solve negative elevation on DTM pertaining to Mavic 2 Pro. I have spent hours looking into and hopefully you could get back to me promptly.

TQVM. 

6

Hi Chin Chye Lim,

Sorry for the delay, investigating took longer than expected.
I would like to point out that you are eligible for personal support which guarantee a first reply within 24 hours. 
You could have submitted a ticket here:
Contact Support

The images that are saved on the drone’s SD card are geotagged by DJI. Lat./Long. coordinates are reliable in the image EXIF, however, there might be some inaccuracies for the altitude depending on the location where you are mapping. We made testing in our office here and we found that the vertical coordinate is off by several meters that can reach an error of 100 meters. Some users noticed the same. Note that this is just an offset meaning that within the model, the accuracy is not affected, only the absolute location is.
As a consequence, we always recommend processing with ground control points (GCPs) in order to fix these uncertainties.

However, if you know your take off altitude you can edit your geolocation file exporting it and sum the take off altitude to the written altitude and importing back:

Hope this will help you to improve your project.

Best

7

I have the same issue last time, GCP would help a lot in correcting Z values. another ways could be using base station and rover for obtaining GCPs (could reach centimeter levels accuracy depend on the GNSS receiver we use)
good luck

8

I agree with Josef. And if an absolute accuracy of a few meters is acceptable for your project, the easiest way to mark GCPs is to use a Web Map Service provider (e.g. Google Earth) and get the values of these GCPs.

Cheers,
Teodora