G'day all,

I have a *.las dataset from which I have derived a canopy image, and a ground image.

I can recall the process to subtract the ground raster from the canopy raster using M8. I'm stuck in 9. Is the process done only in SQL?

thanks

in 9. Is the process done only in SQL?

No, you can do it in a point and click way using the Transform Pane. Images that convey heights are single channel images. In 9 you can add the ground image as a channel to the canopy image (use a raster Join to add a channel from another image, like in this example) and then in the transform pane use the Arithmetic operations between channels on tiles to subtract.

Oh!!!

that is sublimely simple. Thank you.

<edit> as an aside, I connected to a librarylas for the first time. Seamless interaction with a 213GB folder. Really impressive.

You say you derived the two images. Does that mean they were not originally separated and you separated them from the cloud?

You can run interpolation on subsets of the points, typically two classifications. One may have all the points including tree canopy and building roofs, another might only represent bare earth returns. There may be a dozen or more different classifications amongst the point cloud.

That's the case. I separated the ground points (transform select, then copy paste to new table/drawing)and then transform interpolate to create the image. Did the same for the canopy.

Then as per Dimitri's advice above, used join to load the canopy into a second channel.

I'll post a reply with the Arithmetic operation that I used. The value dialogue box has options that are not obvious when working with tiles. (Not obvious = didn't read the very fine manual in absolute detail.)

My question is how you distinguish ground points from low shrubs from water from concrete from buildings from tall trees from clutter. The point clouds I've gotten (already classified) had a complete classification just for random clutter. It would be amazing if you could start with an unclassified cloud and classify them within M9.

My apologies, I should be clear, the dataset contained a classification field.