We are incredibly excited to announce Structure, the first depth sensor for mobile devices!
Structure has been developed by Occipital, the company the Skanect team joined a few months ago, and since then we have been working hard to create software able to run directly on a tablet.
If you want to get involved and help us make this happen faster, we encourage you to back our Kickstarter project. You can get all the details about the new sensor at Structure.io.
For those who will still need a full featured scanning software, Skanect for desktop will continue its evolution, and we will soon add support to stream data directly from the tablet to a powerful computer running Skanect. This will greatly improve both mobility and usability and will enable coloring using the high quality color camera of the iPad.
Being able to scan on a mobile device marks a huge milestone on the road to putting 3D scanning into the hands of everyone and we hope this gets you excited as us 🙂
Here is a post shared by James Allan from Crowtrees Studio on the Skanect forum, showing how higher resolution textures can be obtained using Meshlab and Blender. There are 3 videos showing the process. Many thanks to him for sharing this! Read it on the forum.
Windows users can now enjoy the same new features as the Mac version, check it out!
See our previous blog post related to the Mac release to get the full list of changes.
We are very happy to announce the official release of Skanect 1.5 for Mac OS X Mountain Lion. This new version is optimized to run on recent models of Macbook Pro, Macbook Air, Mac Mini, and Mac Pro.
At this stage all sensors are supported, except Kinect for Windows.
It also includes a number of improvements which will soon be available for Windows too. Here is the full list:
- Export UV texture-mapped models (OBJ and PLY formats).
- VRML export (suitable to upload colored models to Shapeways).
- New export parameters: scale (millimeters, meters or inches) and number of faces.
- Improved user interface.
- Performance improvement (about 15%) for machines without GPU.
- Smoothing level parameter when creating watertight models.
- Track loss detection is now optional.
It also includes several bug fixes, including STL export, more robust file management, and more predictable offline reconstruction.
Many thanks to the early adopters of the Mac beta for their feedback!
To download a copy of Skanect for OS X, click here.
Tagged with: Apple
, OS X
Posted in Uncategorized
Dear Skanect users,
We are very excited to announce that ManCTL Inc., the company behind Skanect, has just been acquired by Occipital Inc, one of the world’s leading developers of computer vision-based mobile apps.
Founded in 2008, Occipital is the creator of RedLaser (acquired by eBay in 2010) and 360 Panorama, the world’s top selling mobile panorama app, downloaded 7 million times.
We will continue to actively develop Skanect within Occipital, and we strongly believe that our new environment will allow us to bring much better 3D scanning to more people on even more platforms. In fact, it already has.
For the time being, not much will change for you, Skanect will be supported as before, the license scheme is unchanged, and the software will continue being maintained and extended.
We would also like to seize the opportunity of this message to thank you all for the incredibly valuable feedback and support you have been giving us so far, which allowed Skanect to keep improving and attract a growing base of users.
You are the ones making it happen, and we are happier than ever to keep on serving you!
The Skanect Team
We are excited to announce a new release of Skanect, which includes many fixes, a reworked scanning feedback, a faster and more accurate CPU fusion and more robust tracking with failure detection and recovery.
Here is the complete list of changes:
- Make the feedback window full-screen during scanning
- Much faster CPU fusion, with improved accuracy
- Tracking failure detection and recovery
- Space bar shortcut to toggle recording
- Pro/Free features clarification
- Fix STL export
- Support for Asus Xtion Pro (depth only)
- Support for non-cubic bounding boxes
- Support for custom intrinsics parameters
- Fix hiccups with some Primesense cameras
- Use local time for generated scan folders
- Fix default color rendering when uploading to Sketchfab
Crowtrees Studio have been scanning a lot of people in events, check out their facebook page to see some great scans!
We are happy to announce a new release of Skanect, which now includes a CPU fallback of the fusion when a compatible graphics card is not available. While slower than GPU fusion, it can still be used in real-time with recent multicore machines, and enables full quality offline reconstruction on all computers!
Here is the complete list of changes:
- CPU fusion
- Support for Kinect for Windows SDK 1.7
- Better automatic calibration of the sensors.
- Minimal distance of 50 cm instead of 80 cm for Xbox Kinect when using the Kinect for Windows SDK.
We’ve just released a bug fix release:
* Colorize should not crash anymore on big scans.
* Exported meshes in the free version should now have proper per-vertex colors.
Thanks for your bug reports!
We are pleased to announce a new update of Skanect, which now supports OpenNI2. This makes it easier to use OpenNI compatible devices, and in particular the new Primesense Carmine 09. With a minimal distance of 30cm, and a sub-millimetric precision at short range, it is the most adapted camera to scan small objects.
Check it out!