3d-houses

Table of Contents

• Introduction
• Folder Structure

Introduction

This project is created with the goal of visualising a building in 3D starting from only the address.

Strategy

1. Use an API call to https://api.basisregisters.vlaanderen.be/v1/adresmatch get a polygon from a house when given an address
2. www.Geopunt.be has a digital height model. With the address we have we find the right files to download. We want to download DTM and DSM data. These are digital height models of flanders which are created by LIDAR. With the DTM and DSM we can create Canopy Height Model, which represent houeses in this case.

An example of DTM, DSM and CHM:

3. Finally we can plot this data to get a virtual representation of the building. Although this is not perfect. Since the data is LIDAR taken from a plane we will only get the top down bird view. As a result, a bridge will look like a wall and an air-balloon will look like a tower when converted to 3D.

An example of a 3D plot (interactive when you run it yourself)

Folder Structure

• Data: All dat will be stored in this folder
  • GeoTIFF: tif files that contain DTM and DSM data, from which we obtain the Canopy Height Model
  • images
    • Saved plots of 3D houess
  • bounding_boxes_coordinates: Pandas dataframe with coordinate pairs as index and column and values as folder number in which the data can be found.
• notebooks: Notebooks to test out code
• utils: module with function we need to build 3D plots -api.py: Functions that use requests to make api calls to https://api.basisregisters.vlaanderen.be/v1/adresmatch to get the polygons of a house/building -functions.py: One function to match a coordinate in the indexes of bounding_boxes_coordinates
• house.py: Defining the house class in which most operations happen, connecting the code base: api calls, plots, ...
• main.py: The file to execute, just enter your address in the correct format and the program will do all the rest.