I can’t share everything.
You have to build your own system.
It begins with reading the pixels off a texture that has noise on it or it is a heightmap texture taken from some place. You need to read it’s pixels.
Next step would be to store the pixel data you got off of the texture to a temporary place and then covert it to usable data for your vectors and T-arrays.
Once you do that you have to load it to some vectors on the axis and offset the values and from vectors load it to an array structure as vertexes.
When you are all done depending on what system you are in inside there is a section that loads all your data to the system you are using.
So you need some form to read the pixels, you decide how. You have to have an index formula for reading them, after that you have to store them to a form of some sort a type, I would use FColor.
You have to find a way to then convert that type to a bit range for your vectors to use, 8,16 etc , 16 bits is usually enough, from Fcolor that is a modified int type to 16 bits for 0 to 1 float range. Then you basicly do data manipulation with vector and arrays, vector logic, etc.
This is just the heightmap system to render a gray scale texture to mesh.
You may want also procedural texturing done to make your own heightmaps.