Hypothetically speaking, if someone managed to create an SDK. What features would you be interested in and what interfaces to access 3d data from the POP would you be interested in?
What toolchains, libraries, etc. would you recommend for creating a 3D point cloud registration, meshing, etc. pipeline.
This is just hypothetically asking… the picture is a hypothetical mockup.
If you are interested to hypothetically work on something like this, let me know.
I am very interested in some SDK.
For my purposes I need create my own pipeline which means
Start Scanning (with selst mode)
Get points of cloud, meshed object with texture
Save into file/memory stream.
It is for headless/userless automatization so I do not need any GUI just call functions and obtain results.
I think it could not be a problem for Revopoint prepare this kind of simple API.
similar needs here.
I could use any information that allows me to create a pcl-point cloud.
I.e. 3D-Points, or depth map.
@p3dt Could you tell us how you did that?
I would love to have an SDK. Just being able to take depth pictures with the POP 2 in my own application would open up so many possibilities.
Could you share code necessary to produce the depth image? I connected to the UVC depth camera buffer, but there is no depth data in it.
The image from the depth camera has a stereo IR pair in R, B and an edge detection in G. There is no depth data.
Feel free to join my discord channel “3d-scanning” to discuss this in detail some time. Might be quicker/easier
@c_v2 that is what the Depth Camera produces . IR based stereo image
it seems that this topic “SDK API” has no recent posts,
however, I believe it’s of high interest, please keep this active.
In my opinion, it could be already a great start to release a simple function :
- load in memory a single frame depth map with or without colors
format it could be compatible with standard ply or others,
and the intent of having easy access from C++, Python and OpenCV libraries and functions.
Thanks a lot
I can understand the interest, but PLY is an export format. We would need the native data that is provided by the scanner to Revo Scan.
Yep, what I mean is that even a raw matrix of points cloud or a depth map would be better than nothing to start working with this development approach and application.
We can’t ask to have a complete SDK soon, but a basic function I believe it’s feasible for Revopoint for being released for experiments.
Eventually not public, but only upon request, if Revopoint want be conservative
I would be very interested in the API. Even a closed one.