FS Trackdraw

FS Trackdraw is a tool designed to help draw driverless racing tracks on top of satellite images (you can also use a white image if you don't need the "real-world context"). The tracks can be exported in CSV format for a 2D simulator. This tool allows you to define the centerline, trackwidth, and the distance between consecutive cones. It also lets you draw a barrier of the available space onto the image.

Feature Ideas

• Randomly generate admissible tracks.
• Load existing CSV-tracks into the satellite image.
• Randomize cone spacings along and normal to the track.

Getting Started

• Left Mouse Button: Defining the Track (Add, Move, or Remove mode).
• Middle Mouse Button: Panning the image.
• Mouse Wheel: Zooming in on the image.
• Right Mouse Button: Defining the Barrier (Add, Move, or Remove mode).

You can follow the steps below to create your track and use it for simulations.

Prerequisites

Before starting, ensure you have the following:

1. Python >=3.10 installed, dependencies in requirements.txt:

   pip install -r requirements.txt

2. A Google Maps screenshot of your location with the scale legend visible. You will need this for drawing the track and scaling.

Step-by-Step Guide

1. Screenshot the Location on Google Maps:

   • Open Google Maps and find the location of the track.
   • Take a screenshot of the area including the scale legend (distance reference).

2. Calculate Scaling:

   • Use the scale legend in the screenshot to calculate the scaling factor from pixels to meters (px/m).
   • This scaling factor will be used to convert the pixel-based track drawing into real-world measurements.

3. Prepare the Location Folder:

   • Create a folder named after your location in the /location_images directory.
   • Inside the folder, place the following files:
     • The satellite image (the original Google Maps screenshot).
     • A config YAML file containing the image file name and the calculated scaling fraction (px/m).

   The folder structure should look like this:

   location_images/
   ├── your_loc/
   │   ├── your_loc_satellite_image.png
   │   └── your_loc_location_name_config.yaml
   

   Example content for the YAML configuration file (your_location_name_config.yaml):

   px_per_m: 10.0  # scaling factor (px/m)
   sat_img_path: 'your_loc_satellite_image.png'

4. Configure the Main Track Configuration:

   • Open the track_config.yaml file located in the config/ directory.
   • Set the standard_location to match the folder name you created in step 3 (e.g., your_loc).
   • You can also adjust the default values for track width, cone distance, and boundary backoff in this configuration file.

   Example configuration (track_config.yaml):

   standard_location: "your_loc"
   track_width: 3.0
   cone_distance: 3.5
   min_boundary_backoff: 10.0
   n_points_midline: 300

5. Run the Track Drawing Tool:

   • Execute the following command to start the track creation tool:

   python src/main.py

   The tool will load the satellite image and display it on a GUI where you can:

   • Add, remove, and move control points that define the track.
   • Zoom in or pan on your image with the middle mouse button.
   • Define the centerline via control points of a cubic spline.
   • Adjust the track width.
   • Using the right mouse button, define an outer barrier as a collection of connected line segments by adding, moving, or removing barrier points.
   • Adjust the cone spacing and boundary backoff interactively.

6. (Optional) Export the Track as a CSV:

   • After defining the track, export the track layout as a CSV file. The CSV contains the cone positions along the left and right boundaries, including tags for the cone color.
   • Save the track layout by using the "Export CSV" button in the GUI.
   • The exported CSV can be visualized with src/plot_csv_track.py.

   The CSV format will look like this:

   tag,x,y
   blue,2.3,5.1
   yellow,2.5,5.2
   blue,2.7,5.3
   

   The coordinates are based on the transformed coordinates from pixels to meters.

Additional Notes

• Track Width and Scaling: The track_width parameter in the configuration file defines the distance between the left and right boundaries of the track. This value is used to generate the track boundaries based on the centerline.

• Boundary Backoff: The min_boundary_backoff specifies the safe distance from barriers. It ensures that the track is drawn at a sufficient distance away from any barrier.

• Conversion Factor: Ensure that the scaling factor is correct and yields reasonable distances in the exported .csv file.

Visualizing the Track

You can visualize the track by plotting the CSV file using the plot_csv_track.py script:

python src/plot_csv_track.py --csv your_exported_track.csv

This will display the track layout and allow you to verify the export.

---

By following these steps, you can easily create a track layout on a satellite image and export it for use in a simulator. The tool offers a straightforward way to define racing tracks while considering barriers defined as polygons and visual obstacles defined via the satellite image.