Sounds like you’re doing some pretty cool analysis!
Are you asking if, given a Cartesian (XYZ) point, you want to get the color of the model at that point? I think that’s going to be a little tricky, if your model is textured, you’d have to read it from the texture, and reverse the transformation from world coordinates to model to texture. I’m assuming you’re trying to read more than one pixel at a time, so I’m wondering if it would be better to just capture the entire scene and then you could read all the pixels directly. Perhaps that would be better to feed into your machine learning algorithm?
Yeah, I’m really doing some analysis by machine learning.
I want to get the Cartesian (XYZ) automatically and simultaneously get the corresponding color at that point.
If I capture the entire scene, I can get the Cartesian(XYZ) and the corresponding color, it will be ok.
The first difficulty I have encountered is I can’t get the data for training my machine learning algorithm.
I feel that the 3D data is more rich than 2D image, so maybe i can apply machine learning algorithm to do something more interesting.
At the first of applying the deeplearning algorithm, I have to get data. So every tool for getting data and transform it to matrix/vector is ok!
I will try the Scene.pick and the discussion, thank you!