PNGetTrackedLandmarks?

In LibPlacenote.cs, there is a method ‘GetTrackedFeatures()’, calling a native method ‘PNGetTrackedLandmarks()’, the sourcode of which is not visible.

Could you explain in a bit more detail, what this ‘landmark’ means? and when/why does it updated? I ask because it looks like it’s updated randomly. Now in our use case, we would like to place/map multiple obejct in the space, and we would like to save an thumbnail for each object. How do I get a thumbnail for each object by using this GetTrackedFeature()?

Hey @clifflinmao,

Landmarks are persistent features we find in the environment. These might change drastically with features appearing and dissapearing over the session as our system decides whether features are dependable persistent features or not.

These are probably not helpful for placing models, your best bet is to place these onto ARKit Anchors (https://developer.apple.com/documentation/arkit/aranchor) while mapping.

OK, Thanks. So from the description above, 'Landmarks’s is the actually ‘anchor’ of placenote?

As the landmark is the ‘anchor’ of placenote, why would we ARKit anchor? As we can place an ‘anchor object’ on that landmark, and then everything other objects we created can have a relative position to that anchor obejct. e.g. In the ‘HelloWorld’ sample, the little dancing man can be the ‘anchor object’, and if we create other objects, their pose can relate to the dancing man. Is my understanding correct?

Hey @clifflinmao,

We have purposely not overwritten ARKit features, we work with ARKit features, not around it. You can continue using ARKit Anchors exactly as they are!

Our landmarks are just that, landmarks, they are not anchors because anchors already exist :).

Hi, Presenjit
May I confirm what you suggest as below:
Currently we are following the Placenote sample ‘HelloWorld’ and ‘StickyNote’ to place multiple objects in the space, then we map and save the map. For all the obejct that we placed, we expect them to be in the accurate position when the map is localized. So far we are just doing as shown in those sample, we are not creating any ARKit anchor for them, because we expect that then the map is localized correctly, all object should appear in the right position. Are you saying that we 'd better not relying on the Placenote Map only, and we 'd better create ARKit anchor for each object?? What would be the difference? With ARKit anchor, those object would appears more accurately/reliably?

Please confirm. Thanks.