I am starting this discussion to bring to together all of the different types of software people are using to process photos collected with their drones and also what they are using it for and what limitations they are finding.
I am currently using the following software:
Agisoft Photoscan Pro
-The data I am running through it was not collected for photogrammetry so I am having some difficulties
-Some images are collected in winter with on the ground making it harder for photoscan to find matching points
-Some images do not have enough overlap or not good enough quality
-Images over forested landscapes sometimes have problems finding matching points
I have using this for quick stitching of images with too little overlap for Photoscan.
I have found it does not work well for long linear set of images
I use google earth to find coordinates for ground control points or georeferencing.
Is an open source GIS software that I use for georeferencing stitched or single images and creating data from the images
I also do a lot of work with LiDAR data and as such am very interested in classifying the point cloud that I can create with Photoscan. The new version has a tool for classifying ground points and then allowing you manually sort the rest. But I am also interested in using the one thing the advantage that photogrammetrically derived points clouds have over raw Lidar data and that is point cloud colours. I am interested in creating a work flow to classify orthos created into feature types (as can already be done) and then assigning these feature type to the point cloud that is also created.
I've been using VisualSfM and CMPMVS. You can run these on Windows, but need 16G memory and an Nvidia card to process the results. With less than 16G memory you need to tweak the cmpmvs file to lower the point cloud data size.
Disadvantages of CMPMVS is that it sometimes leaves holes and in certain cases produces vertices below the mesh surface. This makes normal recalculations trickier.
On ICE... I did a demo on diydrones some time ago showing its results. I think it's pretty poor at performing the actual matching. I reckon it's probably more useful for pano shots without too much rotation or up to 3-4 photo panorama shots:
There are other tools like meshlab, rapidexplorer and cloudcompare to process the point cloud data. Those each have useful functions for density reduction, normal recalculation and so on. I did a demo some time ago on how you can use meshlab and other tools to utilize the areas in visualizations or games:
I use commercial tools for professional work: pix4uav and menci. These produce far better results. You can't use vsfm or cmpmvs commercially without having an agreement with the original authors.
Getting the data georeferenced is more tricky with visualSfM and CMPMVS. Basically you need to figure out where your cameras were. I updated the photo's EXIF (easiest) with each point to get them located correctly.
You can use as well the libraries MicMac developed by the French Cartographic Agency for stitching your images:
I've never used it yet, but I heard that the results are quite good.
Hi Gerard - Does your workflow take into account georeferencing with VisualSfM? For my purposes I need to be able to take measurements from my data and display it on a map with other data.
Hi Arnaud - I have just stumbled accross MicMac and it looks very promising to me.
I would be interested in doing some tests with small (<100 photos) sample set of good data and processing it with several different work flows and having it GIS ready by the end. If anyone has a good data set with lots of different feature types (ex. town, field, forest, water) I would be interested in working with it, provided it has GPS data to go along with it. Up to this point I have only used Agisoft Photoscan Pro and have been very impressed with it, but would like to learn about others experiences with different software.
I'm looking for some aerial pictures to do my test as well.
If someone as some that I could use it would be very nice :)
Guys, do you have some aerial pictures to share :D ?
http://uas.trimble.com/media/downloads (in documentation)
It's possible to georeference through visualsfm, but the process is a bit iffy. You can do this through EXIF gps tags in the photos or gcp's. I didn't try the gcp method yet as I didn't apply them in my session utilizing that.
Utilizing the GPS EXIF method, I'm seeing a precision of 1-2 meters, but that's also related to the process of triggering and recording position used at the time (the APM log can record the trigger position, but this wasn't activated).
Thanks Gerard for the data and the information about visualsfm
MicMac is too complicated to install and to use!
it is a tool of researcher, not user.
In this case, what do you suggest to use ?
I think PhotoScan is the software that offers the best cost / efficiency ratio.
I've read a lot of good comments about this software.
I will give it a try just to see how it works !
Thanks for the information
It's the good way.
You can request a 30 days trial: