Having this mono image, how can I assign a different id to each of its white islands? I tried all conversion options available in the "Mono to Id" operator, but none of them seem to achieve the goal.

https://www.sidefx.com/docs/houdini/nodes/cop/segmentbyconnectivity.html [www.sidefx.com]

This operation breaks the incoming layer into connected components according to a rule.

Parameters
Signature
The layer type that the source accepts.

See Signatures for more information.

Connectivity
This rule determines if pixels are connected or not.

Below Threshold

All pixels below the threshold are connected.

Above Threshold

All pixels above the threshold are connected.

At Threshold

Pixels are connected to their neighbours if they are on the same side of the threshold.

At Levels

Pixels are connected if they are in the same “band” of values. Each band is of size Threshold.

Threshold
The threshold for connectivity. For levels, this is the width of a band of connectivity.

Offset
A global offset for the At Levels mode to adjust where the bands split.

Inputs
source

A Mono or ID layer to detect connectivity in.

Outputs
id

An ID layer with a unique integer greater than or equal to zero for each connected region. Regions that do not match the connectivity region get an ID of -1.

alpha

A Mono layer which is 1 if the pixel met the connectivity rule, zero otherwise.

Perfectomundo! This looks exactly like the operator I needed, Mr. PHENOMDESIGN. Thank you!

I do wonder if this would have been better named "IDbyconnectivity"
I used it about a week ago - and a few days later couldn't find it again :/

Yeah, it could be aliased to this name at least. Last week I used an equivalent of shape splatter node from Substance Designer. But today I forgot what it was called. Perhaps it was Stamp Points? I guess in time, we'll get used to those node names. I might complain about this and that, but I'm very thankful that SideFX finally decided to jump into procedural texturing department to completely dislodge Adobe from their monopoly in this field.

Once sculpting and painting tools improve (and their performance), I might even ditch Substance Painter altogether. I was considering purchasing InstaMAT once its native GNU/Linux version is finally released, but with Copernicus under my disposal, I see no need to do so anymore.

I wonder about this too. Icons are more universal than text. So I think that the icons, well Icons have been proven to be good for memorization and associations, but I think that they help and it is a big part that APEX Graphs are missing. The nodes with Icons are actually the single distinguishing feature from other node software. But then there is also the semantics of meaning etc. but specifically, associating a node with an Icon as opposed to a long feature name is a cognitive affordance.

So the Icon on this references a mesh segmentation and is non isomorphic (does not directly relate to the form of function). I maybe bias but this Colorful Camo kinda looks like a more representative iconography for this.

I am working on semantic embeddings for search and cognitive load lessening. I think that having a "discussion" with the documentation and to question semantically the software about key features is approachable now. This way you can have creative intent that does not have artifacts of the "technology" medium.

I see things moving more inline with Human Machine Teaming Interfaces for Intentional Creative Computing. Machine Programming has some good ideas about this too: