In the example posted in the blog, it seems that it is possible to scan an object and auto-crop the floor and environment it is on? Are there guidelines on how to use Segmentation or other features to enable the auto-crop?
Do you have a more robust version of this, maybe one that integrates Niantic Segmentation, so that it can function in more cases in the wild?
We currently don’t have sample code in that, but feel free to use it together with the semantic segmentation feature in ARDK.
The ground plane cropper works quite well when there is enough ground! Do you also have a contiguous object cropper?
Hi Ina,
Unfortunately we are unable to provide a contiguous object cropper at the moment though it is something that will be available in a future release . At the moment there is no concrete date/ release version but do keep an eye out for new versions of ARDK and check the release notes for information on what features are release in future versions 