Feature Request: Important features for RC photogrammetry workflow

[pypoulp]

Hi, The Reality Capture integration in Houdini is pretty good ! But IMHO it misses some important features.

I guess the SDK is private, so unfortunately it's not possible to help on the coding part, but here are some ideas for improvments :

Alignment Transform

It would be nice to have a node to tweak the alignment, for example, by providing a transform matrix input (to scale / rotate the alignment) or by using a csv flight log as we can do in RC.

rc_alignimages: Oriented BBox Support

I've noticed that if you provide an oriented bbox to the rc_createmodel node, it will recompute a straight bbox internally.

rc_createmodel: additional parameters & Verbose

I don't know if it's exposed in the SDK, but would be nice to have a Houdini progressbar or console log that's more verbose about progress, especially for model reconstruction.

It would also be nice to have a bit more control, especially on image downscale factor before model reconstruction.

RC smooth operation works quite well on large meshes to remove remaining surface noise, would be cool to have an option to enable smoothing after reconstruction.

rc_texture_model: additional parameters & custom mesh support

Texturing node works well but it would be nice to have more control, especially on

• Texturing mode (single texture, fixed texel size, max udim)
• Output format and alpha on/off

The most important feature would be to support custom mesh input in rc_texture_model node, the idea is to do retopo/uv using houdini tools, and the reproject the texture on the cleaned model using rc_texture_model node.

Thanks for all the great work on the Gamedev assets ! Hope it will help.

Best. Paul.