Hello,
I wanted to know if a classification of the dense cloud points can be done in Matic (ground, Buiding, road…) the same as it is done in Mapper??
Thanks,
Thomas Kinsella
Hello,
I wanted to know if a classification of the dense cloud points can be done in Matic (ground, Buiding, road…) the same as it is done in Mapper??
Thanks,
Thomas Kinsella
Hello. Currently, PIX4Dmatic does not have built-in classification. If this is a feature you would like to see included please take a few minutes to post at the link below so that the developers can take your feedback into consideration for future builds.
Latest PIX4Dmatic/PIX4Dmatic Feature Request topics - Pix4D Community
PIX4D survey does have classification features for automatically classifying ground vs nonground and also manual tools for classifying all standard ASPRS classes. Take a look at the link below for more information.
Hi @thomas1, can you tell us more about what you’re trying to achieve? e.g. by showing us an example of project where you need this. This will help to understand the context of your request. Thanks!
If i could fly my drone over an area, walk around and scan areas that are under cover, and throw all the data into a program, add my GCP’s, and have a Point Cloud that I can classify for base earth (like Mapper). Right now you need to pay for every program to do that. Mapper for The drone flight and point cloud classification, Catch for the handheld, Cloud to process the Lidar Data and Matic to see it all as one, but even then you cant classify the point cloud. Please make one app that does all of this and stop making me pay for 4 applications through one company.
@keith3 okay, I believe this can be made easier. I would recommend to use the desktop software PIX4Dmatic for the processing of aerial and terrestrial data with PIX4Dcatch (images+lidar), adding GCPs or Checkpoints, and merging. You can then open this in PIX4Dsurvey and run through the terrain workflow. At each of these steps, you can upload to PIX4Dcloud for sharing the results with someone else if that’s what you’d like to do, check Share to Cloud. We’re working on getting both PIX4Dmatic and PIX4Dsurvey in the same interface, so that you have everything at the same place. This will take us some more time, but both interfaces are already pretty similar. Would be interested in hearing whether this simplifies your workflow.
Even a basic classifier (like Pix4D Mapper) would go a LONG way here.
How have we not got a point cloud classification tool like mapper by now? Mapper used to do it for years now and weve been forced down the Pix4d matic route. This has significantly slowed down our ability to output surfaces because its all very manual now. Im shocked at the delay this has taken to implement given you already have the technology, yet there are updates and changes to UI that to us bear no significance in productivity.
Hi Mitch, what would you use an automatic point cloud classification for? Can you describe what you need to get done? Would an automatic terrain/non-terrai classification be enough or do you need more automatic classes and if yes, which ones?
Can’t speak for Mitch.smith, but we work in an area that has a lot of vegetation / trees.
If we process our flights through 'mapper, we can then filter out a lot of (or all of) the high vegetation points.
Using 'matic now, and not having this option really hurts; Having to manually clean up so much means we lose more time than we would otherwise save using 'matic over 'mapper.
Will this feature be coming soon?
@aus-bd-survey thanks for your input and can you clarify which tool/workflow you apply in PIX4Dmapper? Do you manually assign vegetation points to a point group or do you run the automatic point cloud classification? If you run the automatic point cloud classification, do you find the results are good enough for you, or do you usually do some manual classification afterwards to improve the results?
We use the automatic classification in 'mapper.
It is usually good enough for us to just turn off “high vegetation” and then keep working.
Sometimes it will incorrectly classify things, eg: tops of rock windrows as “high vegetation” meaning we either need to manually re-classify some areas or only filter sub-sections of the data which is easy enough for us.
Do you think this feature will come to 'matic soon?