Back to Projects

Team Name:

AfterRealism


Team Members:


Evidence of Work

Digital Twin

Project Info

Team Name


AfterRealism


Team Members


1 member with an unpublished profile.

Project Description


We wanted to build a system that ingest spatial information at scale and aid us in making decision. In short we want to build skynet :) We started with baby steps; for starter we need a platform that can visualize large datasets. It will be used by diverse group of people, so it has to be a web platform.


#3d #d3 #gis #ml #webgl #python #geojson

Data Story


we experimented with training deep learning model to predict 3D building

we downloaded building footprint data set and merged it with OSM data

we got high resolution aerial imagery and protected it on top of DEM

we used WebGL technology to project large scale datasets in browser


Evidence of Work

Video

Homepage

Team DataSets

Building Footprints City of Hobart

Description of Use This was a fundamental dataset that we used to build our map and plot building 3D buildings. We used it along with OSM to extract building labels.

Data Set

City of Hobart Playground Locations

Description of Use we plotted these on map this was good fun seeing which parks are where along with bike route to plan a weekend outing

Data Set

Urban Art Walls

Description of Use This is such a cool dataset. We ploted places were street art had been done. We wanted to find images online of those street art as well. so you can click on map and see what the painting is like :)

Data Set

Layer: Metro and Tassielink Bus Routes

Description of Use We thought it would be really cool to plot all the routes and try to figure out which residential buildings are furthest away, like show a change of color tone.

Data Set

Hobart Interim Planning Scheme 2015 Zones

Description of Use This was pretty cool dataset. Building footprint on there own don't tell you what kind of building it is. this dataset merged with building footprint tells exactly what the role of a particular building is in the city.

Data Set

Local Overlays - Heritage Areas 2015

Description of Use Initially we thought this would be awesome like places marked in it are landmarks but it identified too places. It was useful in labeling building footprint dataset.

Data Set

DEM

Description of Use We download DEM files from here. But there were in chucks and it looked like a lot of work to stitch them together so for our final project we used the other source mentioned above.

Data Set

contour-0-5m

Description of Use We used this dataset to create a Digital elevation model(DEM). Then we "stitched" high resolution aerial images on top of it to create 3D a model.

Data Set

3D building polygons for City of Hobart

Description of Use We imported it as a restful service in QGIS, and used it in our attempt to train a model to prediction 3D buildings.

Data Set

Challenge Entries

Artificial Intelligence and Machine Learning for a Digital Twin

The City of Hobart is developing a Digital Twin that will contain high resolution photogrammetric data and allow integration with diverse datasets owned publicly or privately. We are keen to explore the use of Artificial Intelligence (AI) and Machine Learning within these datasets.

Eligibility: Use one or more datasets from CoH Open data portal and ensure the submission relates to Artificial Intelligence and Machine Learning for a CoH Digital Twin.

Go to Challenge | 3 teams have entered this challenge.