What hardware/software would you recommend?
Hello Intel,
I’m looking into the different hardware/software solutions for 3D room scale scanning.
Like what’s possible with https://www.intelrealsense.com/3d-scanning-an-environment but for a whole room.
I would like to scan my house for interior design and experiment more in general with this type of technology.
I’m a software developer and run Linux on every device I own. I have a bit of an “open source” mindset.
Also a Blender user and I’m funding its development via the development fund.
Some experience with Meshroom.
What RealSense hardware/software would you recommend? I’m in the market for a new Android tablet and phone so I'm looking into combining your hardware with those.
What’s your (biased) opinion on the Occipital Structure Core hardware/software? Why would your solutions be a better fit for me?
Greetings,
Tim
-
Hi Tim, for this application I would recommend a 400 Series stereo depth camera. It can be used indoors and outdoors and contains Vision Processor D4 hardware that can automatically adapt to changes in the environment. It will therefore be able to cope with the widely varying lighting and shadows that you may find throughout a house.
The D435 and D435i models have a wide field of view and a minimum depth sensing distance of 0.1 meters, enabling you to move the camera near to objects and surfaces for close-range scanning. These camera models will also work with an Android device and you can create your own application for Android via a compatibility 'wrapper' for the RealSense SDK application development software.
As the Intel environment scanning article suggests, Dot3D Pro is an excellent choice of scanning software. Another commercial scanning solution compatible with 400 Series cameras is RecFusion, which could be used with a Windows tablet.
The link below discusses using a depth camera and a T265 Tracking Camera in a pairing together for enhanced reconstruction.
https://support.intelrealsense.com/hc/en-us/community/posts/360038406074/comments/360010057014
Intel do not comment on other companies' camera products, though you are of course free to research and compare different solutions. A key reason why you may consider Intel for your solution though is that it has a single RealSense SDK software development platform that is continually updated over time whilst adding support for new models and retaining compatibility with older models. This enables you to easily update your project to newer camera technology over time, by just changing a few lines of code and also take advantage of new software features and improvements.
-
Hello MartyG,
I bought a D435i, it was delivered quite fast.
I have a follow up question: Why is the RealSense kernel module not available in the mainline kernel? I'm used to Intel hardware having good mainline support.
I'm running Ubuntu 20.04 with the default 5.4 kernel on my desktop and Ubuntu 16.04 on my laptop with a non-default 5.1 kernel (due to flaky hardware).
Neither seem to be supported.According to this I need to verify that a supported kernel version (4.4.0-.., 4.8.0-.., 4.10.0-.. , 4.13.0-..or 4.15.0-.. as of May 2018) is in place.
Are only 4.x kernels supported?In the end I want to create this setup
-
Both Ubuntu 20.04 and Kernel 5.4 work with the RealSense SDK, though neither are officially validated to do so yet. Overall this combination seems to be relatively stable though in the months since the RealSense community began using it.
In the 5.x kernel line, the versions that are officially supported by the RealSense SDK are 5.0 and 5.3. 5.4 works too unofficially.
In the 4.x kernel line, 4.4, 4.8, 4.10, 4.13, 4.15, 4.16 and 4.18 are officially supported.
A RealSense user created and shared a patch to use Kernel 5.1 with the RealSense SDK.
https://github.com/IntelRealSense/librealsense/issues/4271
If you encounter kernel conflicts, there is an SDK installation method called RSUSB that requires an internet connection to carry out the installation but is not reliant on kernel versions or Linux versions and does not require patching.
https://github.com/IntelRealSense/librealsense/issues/6568#issuecomment-642292372
Please sign in to leave a comment.
Comments
3 comments