Support Website Contact Support Blog

Poor result for process image from sequoia camera in pix4dmapper

Dear Pix4d

It seems that I have a problem with the image data from my project by showing as pic below.

I use a camera Installed with Sequoia camera, image overlays is 60% and try to processing RGB in 3Dmap “rapid” mode with Pix4Dmapper 3.1.22.

Then I tried selecting some image covering small areas to run full 3Dmap mode. The result still encountering the same problem

What are the causes and how can I solve?

Thank you for your help

Tony, Can you give screenshot of the flight plan/mapview? We recommend using the ag RGB template for agriculture projects (homogeneous areas) with RGB images

Thank you Momtanu for your reply

Here is my flight plan setup.

It looks like has 2 fight plan, right? Yes. I flew 2 frites, but the settings are the same accept direction. I want to combining together after this,

My work is need the DSM data to create a contour data, so why I trying to use 3Dmap to process. Anyway, that agRGB template can create contour?

Thank you for your help

I will recommend processing these 2 flights as separate projects. You can either

  1. Process all steps and merge in QGIS
  2. Process step 1 of each project, mark MTPs, reoptimise and merge (

Ag RGB template uses alternative calibration method which works better for homogeneous areas than standard calibration (the pipeline that 3D maps template uses). So select ag RGB and then enable all the outputs you need in the processing options.

For sure the price of the seqoia camera (3000 USD ! ) is not justified with the results put out here even though your drone does not have a rolling shutter.

I have test following your recommendation (separate project, Ag RGB template, Select DSM generate as output) : I test in rapid mode reason time save.

The result show it look OK in quality report show

when finish the process, I open it on QGIS the result show as

I don’t know what happen or can you help to setting up the Ag RGB template in detail

Thank you for your help

We have had some cases of holes while down-sampling when rolling shutter correction algorithm is used. Can you process the project with global shutter and check? You can change it in the image editor:

Now, the result look good. RGB map can open like normally.

Anyway, the DSM (DTM) that I select to export is not good enough. Let me guess, it is about the percent of image overlapping that I assigned at 60% is not suitable for create DSM.

Do you have any suggestion about this?

Thank you for your help again

I would say so, we recommend 85% frontal and 75% side overlap for areas like this. Also, parrot recommends sequoia (since it has rolling shutter and the resolution is not as high as other RGB cameras) should not be used for generating DSM. However, if you are using good flight plan, you will get usable results.