{"id":6575,"date":"2023-10-10T04:47:18","date_gmt":"2023-10-10T04:47:18","guid":{"rendered":"https:\/\/handandmachine.org\/classes\/computational_fabrication\/?p=6575"},"modified":"2023-10-10T15:54:36","modified_gmt":"2023-10-10T15:54:36","slug":"jaime-data-physicalization","status":"publish","type":"post","link":"https:\/\/handandmachine.org\/classes\/computational_fabrication\/2023\/10\/10\/jaime-data-physicalization\/","title":{"rendered":"Jaime – Data Physicalization"},"content":{"rendered":"\n

Design Process<\/h2>\n\n\n\n

My work outside of UNM has largely been 3D point clouds, visualization of camera imagery, and mapping. As a result I am familiar with working with elevation data and fire perimeters. For this project I decided to focus on terrain. I liked the idea of exaggerating landscape features to portray a known area in a fun way.<\/p>\n\n\n\n

One of the things that needs to be taken into account when dealing with imagery is radial distortion due to the camera lens. A “fish-eye” lens is a familiar description of a lens with extreme radial distortion, distorting the view on the edges and turning it round in some cases. I was inspired by this “fish-eye” view. I liked the idea of manipulating an elevation tile and exaggerating the curvature of the Earth in that particular spot, making it seem as though the chosen area is a planet of its own.<\/p>\n\n\n\n

I experimented with using radial distortion, which can be achieved by taking the original image coordinate (and scaling it from -0.5 to 0.5), and multiplying it by a polynomial of the radius:<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

xdistorted<\/sub> = xoriginal<\/sub> * (1 + k1<\/sub>r2<\/sup> + k2<\/sub>r4<\/sup> + k3<\/sub>r6<\/sup>)<\/p>\n\n\n\n

Where k1<\/sub>, k2<\/sub>, and k3<\/sub> are radial distortion coefficients. I have the script that takes the array of points from x = [-0.5, 0.5] and y = [-0.5, 0.5] and changes them based on radial distortion, but this really just affects the x and y values of my rectangular array of points. I could have come up with an equation for adjusting the z, but I realized it was easier to manipulate my elevation tile and treat my x and y values like angles and use spherical coordinates.<\/p>\n\n\n\n

Most maps and geological data are in Mercator Projection, which is a type of projection that turns spherical data tiles flat, with their North, South, East, West bounds are Latitude and Longitudinal values. These are angular values with respect to the Earth’s poles and the Prime Meridian. The tiles themselves are a projection of the spherical data to a cylindrical model. The very top and bottom of the Earth (Latitude greater than 85, and less than -85) are technically not included in Mercator Projection because they project to the holes at the top and bottom of the cylinder.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n
\n
\n

Similarly to Mercator Projection, I decided to treat my elevation tiles as though they covered a much larger area of the globe than they actually do. When elevation tiles are at a very high zoom, small area of interest, the x and y coordinates can be treated as though they’re measurements of distance rather than angles. This is the small angle approximation, and basically what we were doing in class with the elevation tile bitmap — treating the bitmap as if the x and y units were in meters or miles or whatever unit of distance rather than an angle.<\/p>\n<\/div>\n\n\n\n

\n
\"\"<\/figure>\n<\/div>\n<\/div>\n\n\n\n

The method I decided to use inaccurately distorts the elevation data. I compute phi and theta from the x and y values of my grid. My new points have a base radius of 10, and I add to that radius based on my the brightness in the bitmap data.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Due to the distortion, trying to wrap the elevation tile area around an entire sphere significantly changes it to the point it becomes nearly unrecognizable. In order to still make the terrain visible, I decided to do a half-sphere. Where phi = [0, 180] (phi is closest to Longitude, in this case). Theta, on the other hand, significantly distorts the elevation when theta is close to 0 and 180 (like the poles of the Earth). Therefore, I limited theta to = [15, 165].<\/p>\n\n\n\n

Next I picked my areas for elevation. I am fond of New Mexico, and wanted areas that I could easily recognize. I picked the Valles Caldera, because I love that area and have enjoyed hiking in it. I also chose the Sandias. I frequently hike trails like La Luz and Piedra Lisa, or drive to the top to hike along the crest. The third area I chose is an island off of Rhode Island called Martha’s Vineyard. My parents have taken me there a few times when they have taught at a workshop there, and I’ve ridden bikes all over the island.<\/p>\n\n\n\n

I used https:\/\/apps.nationalmap.gov\/<\/a> for my elevation data and QGIS<\/a> to transform my images into black and white bitmaps. I originally wanted to include Hawai’i in my dataset objects, but quickly realized that the elevation change in the individual islands can be so extreme that it requires multiple tiles to visualize it. As a result, each bitmap is composed of a different range of grayscale and they would not fit together without further interpreting the data in the geotiff. So I picked regions with a less stark difference in elevation.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n
\"\"<\/figure>\n\n\n\n
\"\"<\/figure>\n\n\n\n
\n
\n

Sandias<\/p>\n<\/div>\n\n\n\n

\n

Martha’s Vineyard<\/p>\n<\/div>\n\n\n\n

\n

Valles Caldera<\/p>\n<\/div>\n<\/div>\n\n\n\n

From the top image, you can see how the elevation begins to significantly distort at the “poles”. Having them reach all the way to 0 and 180 results in really weird geometry.<\/p>\n\n\n\n

I liked the idea of these prints being covers for lights, so I hollowed out the models with a sphere, giving them a smooth inner wall. Initially I had white PLA available, so I printed my first three prints in white. Because I had time, I downloaded the geotiff for Maui in Hawai’i and generated the model.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

I ran out of white PLA, so I got some clear PETG and tried printing the model in that, to see if it would make a better light cover if more translucent.<\/p>\n\n\n\n

Print Results<\/h2>\n\n\n\n
\"\"<\/figure>\n\n\n\n

Sandias<\/h4>\n\n\n\n
\"\"<\/figure>\n\n\n\n

Martha’s Vineyard<\/h4>\n\n\n\n
\"\"<\/figure>\n\n\n\n

Valles Caldera<\/h4>\n\n\n\n
\"\"<\/figure>\n\n\n\n

Maui Print (Transparent PETG)<\/h4>\n\n\n\n
\"\"<\/figure>\n\n\n\n
\"\"<\/figure>\n\n\n\n

Reflection<\/h2>\n\n\n\n

Despite the fact that the white PLA is less transparent than the clear PETG, the material better showcases the elevation. I like my PLA prints better than the clear PETG.<\/p>\n\n\n\n

Initially I had considered including fire perimeters in the prints, given my background in wildfire and emergency management. I would have included the data if I had more time. If I was going to include fire perimeters I wanted to make sure that whatever art piece I made was not somehow disrespectful to the individuals who lost their homes and lives due to these emergencies. That was part of the idea behind the design in the beginning, to create a piece that allowed the person viewing it to better understand the extent of a wildfire, and to be able to hold it in their hands.<\/p>\n\n\n\n

JaimeGouldLA3.gh<\/a>Download<\/a><\/div>\n","protected":false},"excerpt":{"rendered":"

Design Process My work outside of UNM has largely been 3D point clouds, visualization of camera imagery, and mapping. As a result I am familiar with working with elevation data and fire perimeters. For this project I decided to focus on terrain. I liked the idea of exaggerating landscape features to portray a known area in a fun way. One […]<\/p>\n","protected":false},"author":30,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[37,24],"tags":[],"class_list":["post-6575","post","type-post","status-publish","format-standard","hentry","category-data_phys_23","category-studentwork23"],"_links":{"self":[{"href":"https:\/\/handandmachine.org\/classes\/computational_fabrication\/wp-json\/wp\/v2\/posts\/6575","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/handandmachine.org\/classes\/computational_fabrication\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/handandmachine.org\/classes\/computational_fabrication\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/handandmachine.org\/classes\/computational_fabrication\/wp-json\/wp\/v2\/users\/30"}],"replies":[{"embeddable":true,"href":"https:\/\/handandmachine.org\/classes\/computational_fabrication\/wp-json\/wp\/v2\/comments?post=6575"}],"version-history":[{"count":4,"href":"https:\/\/handandmachine.org\/classes\/computational_fabrication\/wp-json\/wp\/v2\/posts\/6575\/revisions"}],"predecessor-version":[{"id":6912,"href":"https:\/\/handandmachine.org\/classes\/computational_fabrication\/wp-json\/wp\/v2\/posts\/6575\/revisions\/6912"}],"wp:attachment":[{"href":"https:\/\/handandmachine.org\/classes\/computational_fabrication\/wp-json\/wp\/v2\/media?parent=6575"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/handandmachine.org\/classes\/computational_fabrication\/wp-json\/wp\/v2\/categories?post=6575"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/handandmachine.org\/classes\/computational_fabrication\/wp-json\/wp\/v2\/tags?post=6575"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}