What image-set problems do you wish you could catch before processing?

PIX4D Café
1

Hi everyone,

I’m building a tool called Imagery Workbench to help identify potential problems in drone-mapping image sets before they’re uploaded or processed.

The idea came from a familiar frustration: issues such as blurry images, exposure changes, weak GPS metadata, inconsistent altitude, sequence gaps, or incomplete coverage may only become
obvious after processing has already consumed considerable time.

I’d like to understand how people currently handle this across PIX4D and similar workflows:

  • What image-set problems do you most wish you could catch before processing?
  • Which checks do you perform immediately after a flight?
  • Have any particular checks reliably prevented a failed reconstruction or a return visit?

I’m especially interested in whether coverage is usually checked visually or against a planned boundary or polygon.

This is workflow research for a tool I’m actively building, rather than a product announcement. I’d appreciate any examples from real projects—even cases where the warning signs were only
obvious in hindsight.

2

Hi,

  • Dewarping: having a DJI P4R, each time I switch batteries it comes back to the default “Dewarping - ON” mode, so sometimes I will end with at least 2 groups: the Dewarped set and the Warped one. This info is stored inside the image as drone-dji:DewarpFlag=“0” (Dewarping - OFF) or drone-dji:DewarpFlag=“1”(Dewarping - ON), so I think it could be easy accessible to a “Photo-checker”: in our case, we try at first to check it visually (rounded corners near the border, fisheye effect, etc), and in case of doubt, we randomly choose a couple of photos to open them with a text editor