Just recently I took part in generating a series of 5 ortho maps down a Gravel River system to look at Gravel migration and measuring areas of gravel beaches and working out cut and fill.
1st question. Can you stitch all 5 maps together to create on seamless map or would this be too big a file.
The local Regional Council I work for here in NZ is looking to put the maps into their GIS system and the River Engineers are interested in getting me to map before and after floods to workout the movement of gravel down the system.
2nd Question. By chance when I flew these maps the river was very clear and playing around with the Distance Annotation and placing a marker across a deep pool in the river I found the height profile it displayed looked to be showing the bed level. I am yet to go back and measure the profile of the pool with our ADCP instrument but it looks to be very similar. Is it possible that in very clear water the drone images are good enough for DD to actually pick up the riverbed level, if so this could be a very valuable tool for us Hydrologists.
Any thoughts would be much appreciated
Hope you are doing well and thanks for swinging by the DroneDeploy Forum. I’m happy to address your questions below:
If you are interested in merging several maps, The best course of action here is for you to upload all the images to a new map using the “Upload images” button that is located in the blue + icon at the bottom of your dashboard. That way, you can upload all the images to one single map. Please note that this action only allows uploading the maximum number of images that your subscription level allows. Otherwise, please note that you can export your maps as GeoTIFF and use a GIS software to merge them. You can use this tutorial to carry that out process using QGIS.
As for the reconstruction of the river, this issue seems to be related to the nature to stitch water itself. Water and generally anything with a highly reflective surface are problematic for any photogrammetric stitching solution. Photogrammetry works by matching keypoints across multiple photos. Water presents multiple challenges because it is constantly physically moving slightly between pictures and it is highly reflective. Reflectivity causes specular highlights that also change location between photos. Together, it makes finding keypoints nearly impossible. Therefore, stitching water or any highly reflective surface typically results in spiky, glitchy artifacts in the digital elevation model and sometimes holes in the orthomosaic.
I hope you find this information helpful. Happy mapping!
Thanks for the reply. I will try to reprocess all images but I think the number will be over my pro plans quota. I am waiting for the surveyed GCP’s to help with accuracy. What would be a fair estimate on height accuracy with the current maps I have had processed without GCP’s. Is it .500m or better.
In terms of the distance annotation when you draw a line across the water surface it is picking up a shape similar to the riverbed shape which is why I asked if the software is able to read elevation thru the water, your explanation makes sense though about distortion from shiny or reflective surfaces. Maybe it is a coincidence.
If you are looking to add GCPs to your maps, you would have to upgrade your account to our Business subscription. If you are in our Business plan, you can upload up to 3000 images per map, and you could then upload all the images that you want to merge together.
As for the accuracy of your non-GCP maps, it depends on several factors such as the accuracy of your drone’s GPS, your flight parameters, the atmospheric conditions, etc. I highly recommend you to take a look at our How Accurate is My Map? support guide to learn more about the accuracy of your DroneDeploy maps.