Pointcloud improvement #746
Conversation
Since .bin and .cin have different repartitions of the points inside a node, they must use different algorithms. For .cin we want to draw a %-age of each node because points are evenly distributed. For .bin, we sort the nodes by 'importance', and stop drawing nodes when we have enoug points.
package-lock.json
Outdated
| @@ -5954,7 +5034,7 @@ | |||
| }, | |||
| "onetime": { | |||
| "version": "1.1.0", | |||
| "resolved": "https://registry.npmjs.org/onetime/-/onetime-1.1.0.tgz", | |||
| "resolved": "http://registry.npmjs.org/onetime/-/onetime-1.1.0.tgz", | |||
There was a problem hiding this comment.
No idea where this change comes from.
There was a problem hiding this comment.
Can you try to run again npm i in your branch to see if package-lock.json changes?
There was a problem hiding this comment.
nope, doesn't change
There was a problem hiding this comment.
I've pushed a commit with a package-lock.json generated with npm 6. Can you try if your npm i still works ok? if so, we can use this one I guess.
There was a problem hiding this comment.
It works, but npm i reverts all your changes to package-lock.json, eg:
- "find-up": "^1.0.0",
- "read-pkg": "^1.0.0"
+ "find-up": "1.1.2",
+ "read-pkg": "1.1.0"I guess it's expected since I don't use npm 6...
| } | ||
| } else if (!elt.promise) { | ||
| const distance = Math.max(0.001, bbox.distanceToPoint(context.camera.camera3D.position)); | ||
| // Increase priority of nearest node |
There was a problem hiding this comment.
I don't understand the "nearest" here.
There was a problem hiding this comment.
nearest = closest to the camera.
There was a problem hiding this comment.
Oh! right... I find the comment more confusing than the code alone :-)
| } else if (!elt.promise) { | ||
| const distance = Math.max(0.001, bbox.distanceToPoint(context.camera.camera3D.position)); | ||
| // Increase priority of nearest node | ||
| const priority = computeScreenSpaceError(context, layer, elt, distance) / distance; |
There was a problem hiding this comment.
Can't you set elt.sse here? You'd have it already on line 217 this way.
There was a problem hiding this comment.
I could but that would complexify a bit the code at line 127 (with a if (elt.sse is already computed)) for something that only happens once per node: when we decide to fetch them.
There was a problem hiding this comment.
Would it be guaranteed to be computed on line 127? (because elt.promise would be undefined the first time update is called on this node). But even if so, I agree this is a bit fragile. We can keep it like that.
Project average point spacing on screen and compare to sse threshold. It's a simplification of the approach proposed in #674. This allows to drop a dependency (convexhull) and brings the SSE compuation for pointcloud closer to what is done for other geometries.
peppsac
force-pushed
the
pointcloud_improvment
branch
from
131bd25 to
89b6cbf
Compare
|
thanks! |
(reopening #732 )
2 commits improving pointcloud support for itowns.
The first one fixes the point budget handling for .bin format.
The second reworks SSE: it's similar to #674 but simpler and only for pointclouds.
Full commits message:
Edit:
The SSE value computed here is the estimation of the maximum distance between 2 points in a given node.
PotreeConverter guarantees that the distance between each pair of points is >=
metadata.spacing / node.levelin all nodes. This is easy to visualize if you display successive children (using the 'sticky' debug tool):The SSE is the estimation of this distance projected on the screen.