Formatting high-resolution images for useful visibility in Survey123 reports

[MattEdrich]

Hello everyone,

We are trying to develop a secondary version of our field workflow that acquires imagery entirely from technician-operated drones. Our primary workflow is predicated on technicians using their smart phones to capture certain images alongside feature attributes, and the report templates I have created have so far been based on smart phones being the pathway through which we conduct our surveys.

I have no issues with submitting forms using drone imagery through the Survey123 web app, but when I generate reports with those images, the images are compressed enough that they aren't usable in the reports. I have found, however, that deleting the image from the generated report and replacing it with the same image imported directly from the drone's SD card results in acceptable image quality after it has been resized to fit in a typical .docx. I don't understand why the image versions in the generated report become unusable, nor why replacing them with resized raw images works.

The native resolution I am working with in this case is 8192 x 5460, clocking in generally around 8MB in size, though some can be larger than 10MB. Like I said above, our reports are formatted to display images taken with 12MP smart phone cameras (4032 x 3024 pixels), which I have generating using ${img | size:544:306} - the specific numbers for sizing were settled upon with feedback. When I first made these templates, I found that simply using ${img | getValue:""} put in images that were far too large to be viewed in a .docx. I haven't been able to successfully wield ${img | getValue:"" | size:544:306} to insert what I thought would be native-resolution images resized to fit in a .docx.

It is entirely possible that I don't actually understand digital image dimensions and processing like I thought I did. Everything I have tried to automate the reports using drone imagery has come up short. Our surveys can sometimes contain upwards of 100 images, so it would amazing to not have to QA every image in them to the degree that we currently have to. I don't really know much about photography, but I suspect that my issue is at least related to the different image qualities generated by a 12 MP camera vs a 20 MP one. I have not made any changes to the maximum image size that can be submitted, either at the form level or the question level.

If anyone can chime in with an explanation of the nuances of image submission/reporting, and/or how I might make functional use of the drone imagery, I would be extremely grateful!

Resources consulted prior to posting:
Media questions—ArcGIS Survey123 | Documentation
Optimizing Images for Survey Forms - Esri Community
How To: Improve image quality in ArcGIS Survey123 feature reports (esri.com)

Thanks!!