Replies: 2 comments 1 reply
-
Current logic preserves attributes named _ID, _BATCH_ID or _FEATURE_ID_n, as they are known to be integers semantically. I'm assuming the attributes you are using aren't named that? What's their accessor type in the glTF file? |
Beta Was this translation helpful? Give feedback.
-
The attributes are named _CLASSIFICATION and _INTENSITY and the accessor type for these attributes is scalar. I was wary of the The attributes are included using the This is from the glTF file {
"EXT_structural_metadata": {
"schema": {
"id": "pts_schema",
"name": "pts_schema",
"description": "point cloud point attribute schema",
"version": "1.0.0",
"classes": {
"point": {
"name": "point",
"description": "Properties of point cloud points",
"properties": {
"INTENSITY": {
"description": "Laser intensity",
"type": "SCALAR",
"componentType": "UINT16",
"required": true
},
"CLASSIFICATION": {
"description": "Point classification",
"type": "SCALAR",
"componentType": "UINT16",
"required": true
}
}
}
}
},
...
"meshes": [
{
"name": "PointCloud",
"primitives": [
{
"extensions": {
"EXT_structural_metadata": {
"propertyAttributes": [0]
}
},
"attributes": {
"COLOR_0": 1,
"POSITION": 0,
"_CLASSIFICATION": 3,
"_INTENSITY": 2
},
"mode": 0
}
]
}
]
}
} |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
I'm using meshoptimizer trough gltfpack with GLTF point clouds in a 3DTiles 1.1 format. The optimizer drastically improves the size and unpacking speed, but some key integer fields, most noticebaly Classification and Intensity are stripped. Classification is a uint8 and intensity uint16 which get stripped out during compression. LAS specification
I saw the discussion in #529 about not supporting arbitrary custom data, but is it possible to add support for a generic integer attributes so fields like this can be kept?
Beta Was this translation helpful? Give feedback.
All reactions