====== Topographic Lithophane ======

===== About =====

This project is meant to create a decorative map of the US. A typical lithophane of a height-map creates a relief map. The problem with this is that the scale of the relief is determined by the thickness of the plastic necessary to create the grayscale. This happens because the back-plane of the lithophane is flat. What sets this project apart is creating a back-plane with depth so that the scale of the relief can be controlled. This involves creating two separate point clouds from the same dataset and stitching them together into a single stl. This creates an inner and outer profile that are very similar, but with a shell thickness that creates the lithophane effect. Combining both the lithophane effect and an exaggerated relief effect should create a visually impressive piece that looks good both lit and unlit.

===== Progress =====

To start I downloaded the Digital Elevation Model (DEM) of the entire US. I am specifically using the 1/3 arc second (roughly 10m) resolution dataset from the USGS. This was several hundreds of gigabytes of data as grayscale TIFF images or roughly 9.28x10<sup>10</sup> pixles. These images were imported in QGIS to make sure that they all had the same reference height and height scaling. State boundaries were imported and the TIFFs were clipped to create individual state files. Based on the size of the resulting map, these state images were downsampled to approximately match the resolution I could print them at to shrink the sizes of the files I needed to process.

I am currently writing a python script to convert the DEM data into STL files.

===== Python Script =====

The python script takes five inputs: the tiff file, the desired relief scale, the image scale, the material thickness that represents the minimum image value, and the material thickness that represents the maximum image value.

To create this script, I read through the source code for [[https://github.com/cvr/dem2stl|dem2stl]] and [[https://github.com/anoved/phstl|phstl]] which both take DEM data and turn it into an STL but don't do any of the layer stitching I need. They were a great starting point that saved me a lot of time.

I am using the [[https://github.com/thearn/stl_tools|stl tool library]] to simplify creating the STL.