When doing simple annotations like sky removal (white sky noise is a big problem on towers) the longest part is waiting on Pix4Dmapper to segment each image. It would be a great improvement to add a command and/or option in Processing Options to automatically segment each image during Step 1 so the labor hours of annotation can be reduced.
Noted! In my understanding, the pain point here is that one needs to wait for the segmentation of an image while doing annotations, which adds time to the manual labor a person is doing. Please correct me if I misunderstood. It is always good to understand the root problem, because there are generally several ways to address it in practice.
The suggestion of running the segmentation on all images during step 1 suggests that you would be willing to have a much longer processing time for the benefit of reducing manual labor time, is this correct?
Thanks for your feedback!
I think if in the process menu have a option for segment all images, it would reduce the waiting time to using annotations.
100% correct Pierangelo. For my business it is much more costly for labor hours vs. Pix4D hours. I don’t need this every time so an optional selection is needed within Step 1 or even independently.
Thanks for the confirmation Adam and to your input Michel! I reported the request to our development team.
I just fully agree with Adam. The main pain while annotating images is the repeated initial waiting time needed for the image segmentation. As it cannot be done in batch or automatically in a preprocessing step, it just cost a lot of interactiv working (or better said waiting) time.
Both suggestions (optional segmentation during initial processing in step 1 or started independently for all images) would be a great improvement. The second option with an independent processing seems a more flexible option to me. It would be even better, if one could select the images to be segmented and start a batch processing for the selected images, directly from Pix4D or from an external utility application.