AgriFieldGenerator

AFG is a python application that allows you to generate large agricultural landscapes for Bohemia's Enfusion engine with little effort. It is based on Voronoi tessellation.

1. Dependencies
2. Installation
3. Usage
4. Step by step guide
5. Changelog
6. Backlog

1. Dependencies

See or use requirements.txt

2. Installation

a. Clone this repository :

git clone https://github.com/Tanin69/AgriFieldGenerator.git

b. Navigate to the project directory :

cd AgriFieldGenerator

c. Install the required Python libraries :

pip install -r requirements.txt

3. Usage

You can run the AgriFieldGenerator with various command line arguments. For example :

python run.py --points --generator random --voronoi

This command will generate points using a random generator and then generate a Voronoi diagram based on these points.

usage: run.py [-h] [-s] [-po] [-pt] [-g {random,grid,rectangle,rect_tiling}] [-v] [-pp [[0-1]]] [-c] [-m] [-me] [-pl] [-a] [-d {main_polygon,seed_points,voronoi}]

Run the AgriFieldGenerator.

options:
  -h, --help            show this help message and exit
  -s, --svg             Generates a svg file from an Enfusion layer file containing spline entities.
  -po, --polygon        Generates the main polygon from svg file.
  -pt, --points         Generates points schema.
  -g {random,grid,rectangle}, --generator {random,grid,rectangle}
                        Choose the type of point generator.
  -v, --voronoi         Generates the Voronoi diagram.
  -pp [[0-1]], --pp_curve [[0-1]]
                        Curves some random borders. If passed without a value, defaults to 0.5.
  -c, --colorer         Generates the colored polygons.
  -m, --mask            Generates the masks.
  -me, --merge          Merge the masks with Enfusion surface texture masks.
  -pl, --polyline       Generate polylines between polygons.
  -a, --all             Run all the processors.
  -d {main_polygon,seed_points,voronoi}, --display {main_polygon,seed_points,voronoi}
                        Display the results of a given processor.

4. Step by step guide

a. Create your work directory structure

Somewhere on your computer, create the following structure:

• work/
  • ProjectName/
    • sources/
    • file.png

b. Edit the config.json.example file

{
    "work_dir": "/path/to/your/workdir/", <- Path to your work dir
    "project_name": "YourProjectName",    <- Well... Your project name, named like your ProjectName directory
    "source_files": {
        "svg_filename": "output.svg", <- Name of the svg file generated from the Enfusion Spline
        "svg_height": 16257, <- height of your svg in pixels, must be the same as your satmap file and your terrain in Enfusion
        "svg_width": 16257,  <- idem, but for the width
        "tile_size": 512     <- tile size in Enfusion in pixels
    },
    "enfusion_texture_masks": {
        "etm_path": "/path/to/your/enfusion/texture/surface/masks", <- Surface texture mask. They must of course have been exported previously via the Enfusion Workbench
        "etm_1": "Crop_Field_01.png", <- First surface mask. Maximum is 4 surface masks. 
        "etm_2": "Crop_Field_02.png", <- Second surface mask
        "etm_3": "ZI_Crop_Field_03.png", <- Third surface mask
        "etm_4": "Grass_02.png" <- Fourth and last surface mask
    },
    "enfusion_surface_map_resolution": 1.000795, <- Surface map resolution as indicated in Enfusion (see terrain tool, info & diag panel, suface mask section, Resolution)
    "palette": ["#89723e", "#4f4333", "#7e895e", "#3a4422"], <- RGB color code for the preview. The first color is for the first surface texture and so on
    "borders": {
        "min_border_width":0.1, <- borders between fields. Randomly generated between min and max
        "max_border_width":5
    },
    "point_generators":{ <- point generators parameters. Experiment with them ;-)
        "random": {
            "num_points": 50
        },
        "grid": {
                "nx":10,
                "ny":10,
                "rand_offset_x":5,
                "rand_offset_y":5,
                "rand_step_x":2,
                "rand_step_y":5,
                "angle":240
        },
        "rectangle": {
                "num_rectangles":10,
                "min_width":1,
                "max_width":5,
                "min_height":1,
                "max_height":5
        }
    },
    "paths": {
        "__comment": "Don't modify anything here, unless you know what you are doing !",
        "source_dir": "sources/",
        "save_dir": "saves/",
        "save_data_dir": "data/"
    }
}

Save this file as config.json

d. Launch the script (cf. usage)

For example :

python run.py -a

e. Import your texture masks in Enfusion, and follow the Enfusion process

I've made a tutorial for this (painful) part : see https://docs.google.com/document/d/1Ofb3NplPc76hag4b1zzj7Z689JoD68kPY21FDXRLEo4/edit?usp=sharing

5. Changelog

• 1.2.0, 2024/06/23

  • Convert Enfusion splines to SVG (new arg to generate svg : -s)
  • Add curvatures to random borders (new arg to generate curved borders : -pp)
  • Refactored display method, now called from super class (new arg : -d )
  • Cleaned imports and updated requirements.txt

• 1.1.0, 2024/05/27 :

  • Generate polylines on borders in Enfusion format
  • List of affected tiles after masks import in Enfusion terrain exported to save_dir/polygon_tiles.txt

• 1.0.0, 2024/05/20 : first release

6. Backlog

• Use the satellite map as a background image for preview.png
• Add a display method to show Enfusion tiles on the polygon
• Error handling for a potential geometry error when processing svg into polygon
• Take into account the terrain height variations for the generation of points