lego-art-map-generator

Scripts for generating custom mosaics for the LEGO 31203 World Map set.

The mosaics aim to imitate the same style as the original 'World Map' set, with white tiles representing land and coloured tiles representing the bathymetry.

This project is not affiliated with The Lego Group.

Examples

Below is a proof of concept that show a recreation of the LEGO World Map using the scripts in this repository. One can tweak the parameters to get different results, so this is just an example of what is possible using these scripts.

Recreation	Original (By LEGO)

The actual purpose of this repository was to create custom maps. The land and sea-depth images hopefully provide some context as to how the bathymetric data is used to distribute the coloured tiles.

Rendering	Land	Sea-depth	Result
World Map
North Sea
Iceland
Denmark
Greece
Madagascar
New Guinea
Corsica
New Zealand
Scotland
Scotland (2)
Scotland (Mainland)

Project structure

The project is grouped in three sections:

• data - Where all the bathymetry and coastline data is stored.
• map_generator - Scripts to generate the mosaics
• map_analysis - Tool scripts to analyse tile distributions and bathymetric data.

Pre-requisites

First make sure python is installed. Then run the following commands to install dependencies and download map data.

make install
make download_ne_data
make download_gebco_data

Map Generation

Map generation is split into 5 steps, each is its own script:

• Step 1: Generate a grayscale image of the land-masses.
• Step 2: Convert the image in step 1 into a 1-bit image.
• Step 3: Modify the image in step 2 by adding a shadow to the edges of each landmass.
• Step 4: Generate a grayscale bathymetric map.
• Step 5: Use the images in step 3 and 4, to distribute the coloured tiles according to proportions outlined in a CSV.

Step 1	Step 2	Step 3	Step 4	Step 5

The Makefile has some examples of how the scripts are used in practice.

Step 1 - Coastline generation

This step comes in two varieties.

To recreate the world map similar to the LEGO world map, run this script:

poetry run python map_generator/step_1_land_grayscale_world_map.py step_1.png

To generate a custom map of a specific part of the globe, run this other script, which takes in numerous parameters for controlling the projection. This includes center (longitude/latitude), scale (meters per pixel), rotation (Degrees). The UTM projection is used, and will adapt to the center position you pick to minimize distortion. The size of the canvas can also be controlled using size.

poetry run python map_generator/step_1_land_grayscale_utm_map.py step_1.png --size=80,128 --center=1.8,58.8 --scale=15000 --rotation=25

Here is an example of the output:

Step 2 - Converting the image to a 1-bit image

This script will convert a grayscale image of the landmasses into a 1-bit image consisting only of black and white pixels. The algorithm used can be controlled with the mode flag.

poetry run python map_generator/step_2_grayscale_to_1bit.py step_1.png step_2.png --mode=custom_1

The different modes are:

• threshold uses a simple 50/50 Threshold Filter,
• dither uses Floyd-Steinberg Dithering,
• custom_1 uses a custom algorithm I made that does some vague kernel filtering that takes into account the neighboring pixels. This mode is recommended because the other modes often lose detail or add unwanted details.

threshold	dither	custom_1

Step 3 - Add a shadow to the landmasses

This script takes any black and white image from step 2 (step_2.png), adds a dark-blue shadow to the white tiles on the right-hand side, and saves it to an output path (step_3.png)

poetry run python map_generator/step_3_land_shadow.py step_2.png step_3.png

Step 4 - Bathymetry generation

Warning: This script might take a while to run.

This script generates a grayscale image of the bathymetry data. The grayscale image represents sea-level as #FFFFFF, and the lowest sea-depth as #000000. Similar to Step 1, there are two scripts to choose from.

To recreate the world map similar to the LEGO world map, run this script:

poetry run python map_generator/step_4_sea_grayscale_world_map.py step_3.png step_4.png

To generate a custom map of a specific part of the globe, run this other script, which takes in similar parameters to the script in Step 1. An additional argument is required for determining the lowest sea-depth using max-depth, which is measured in meters.

poetry run python map_generator/step_4_sea_grayscale_utm_map.py step_3.png step_4.png --size=80,128 --center=1.8,58.8 --scale=15000 --rotation=25 --max-depth=3500

The first argument of both scripts is an image that is used as a mask for determining which pixels to calculate a sea-depth for. For example, if a pixel already has a colour assigned against it (e.g. #FFFFFF) it will skip on to the next pixel. Only when the colour is #000000 will the script compute a colour.

Step 5 - Generate the final mosaic

The final and most complicated script will take in multiple input files to distribute the coloured tiles on-top of the image generated in Step 3.

The parameters, in order, are:

• The image from Step 3.
• The image from Step 4.
• A CSV of the proportions, which correlates image brightness to colour distribution. Each column in the CSV represents the proportion that each colour has at a particular "depth" in the map. The headers represent the brightness values of the image from 0 to 255 in the image from Step 4.
• A CSV of the tile counts for each colour that can be used in the final mosaic.
• The output file path.

poetry run python map_generator/step_5_sea.py step_3.png step_4.png data/north_sea_map_brightness_tile_proportion.csv data/world_map_max_tile_counts.csv step_5.png

Step 6 - Generate High-res image (Optional)

To create an enlarged scale of the final image with rounded tiles instead of basic pixels, run this script.

poetry run python map_generator/step_6_pixels_to_lego.py step_5.png step_6.png

Map Analysis

count_tiles_from_image.py

After generating the 1-bit image in Step 2 you might want to check that you have enough white tiles to actually create the image. This script will be helpful in doing a quick count of all the pixels in an image.

Below is a quick example of what the script output looks like:

poetry run python map_analysis/count_tiles_from_image.py step_3.png
0,0,0:          6832
01 - White:     3039
02 - Navy:       369
total:         10240

count_tiles_from_ascii.py

When I started this project, I created a bunch of ASCII Grid files that represented the tile placements in the official LEGO World Map. You can see them in /data/lego_world_map_ascii/.

I created a script that counts the unique numbers in the files, similar to the count_tiles_from_image.py script.

poetry run python map_analysis/count_tiles_from_ascii.py data/lego_world_map_ascii/*

The other scripts

Two other scripts exist in this directory, but they are mainly written in a rush to do some quick analysis of the data. I may or may not document this in the future.

Tile counts

The official box set contains an excess number of tiles to allow you to customise the world a bit, but the there are different amounts for each color.

The total number of tiles for each color are listed below:

Tile color	# of tiles required for World Map	# of tiles in box set according to booklet	Actual # of tiles in my box
	3062	3064	3198
	392	393	410
	1606	1607	1668
	1878	1879	1956
	529	601	621
	1019	1060	1106
	724	725	752
	598	599	619
	229	601	625
	203	601	627
Total	10240	11130	11582

I also collected statistics on individual bags that I found in my set. Your tile count will probably be different to mine, since there appears to be some randomness.

Bag name	tile count
White bag 1	1065
White bag 2	1067
White bag 3	1064
White extras bag	2
Navy bag 1	408
Navy extras bag	2
Cyan bag 1	829
Cyan bag 2	837
Cyan extras bag	2
Teal bag 1	977
Teal bag 2	977
Teal extras bag	2
Green bag	619
Green extras bag	2
Olive bag	1104
Olive extras bag	2
Beige bag	750
Beige extras bag	2
Yellow bag	617
Yellow extras bag	2
Orange bag	623
Orange extras bag	2
Coral bag	625
Coral extras bag	2