Baseball Outfield Distance Estimation

This program (outfield_estimation.py) reads a ballpark image ("image-mode") or downloads it via Google Maps API ("map-mode"), and computes distances from home plate to the outfield wall at 91 evenly spaced angles between the foul lines.

Getting Images via Google Maps

To obtain historical images (if current Google Maps imagery for a certain ballpark is insufficient, e.g. park is roofed), see this short instructional video. Between Steps 2 and 3, select the clock icon at the top to select the year you want the image from.

Prerequisites

• Python 3
• Google Maps Static API key

Required Packages

pip install opencv-python numpy matplotlib requests

Execution

These are bare-minimum execution commands: review Optional Arguments for further customization.

Image-Mode:

python outfield_estimation.py --image-path <YOUR_IMAGE_PATH>

Map-Mode:

python outfield_estimation.py --latitude <BALLPARK_LATITUDE> --longitude <BALLPARK_LONGITUDE> --api-key <GOOGLE_MAPS_API_KEY> --image-path <IMAGE_SAVE_PATH>

(Whether program runs in image-mode or map-mode depends simply on whether coordinates (and API-Key) are provided).

Required Arguments

--image-path: If (--latitude, --longitude, --api-key) specified, downloads image via API to this path; otherwise, assumes user has image already downloaded, and reads from this path

Required Arguments (Map-Mode Only)

--latitude: latitude of desired ballpark
--longitude: longitude of desired ballpark
--api-key: Google Maps API Key

Additional Optional Arguments

--csv-path: Filename to which to save computed distances (NONE if no save desired, default distances.csv)
--vis-path: Filename to which to save distance visualization (NONE if no save desired, default distances_vis.png)
--graph-path: Filename to which to save angle/distance graph (NONE if no save desired, default distances_graph.png)
--field-type: Field type (MLB, MiLB, Olympic), defines warning track length (default MLB)
--min-distance-feet: Minimum distance (in feet) to consider (default 275)
--smooth-level : Maximum allowed jump (in feet) between consecutive measurements (0 if no smoothing desired, default 5)

Additional Optional Arguments (Map-Mode Only)

--zoom: Google Maps zoom level (default 19)
--size: Google Maps image size (default 600)

Output Interpretation

• Distance CSV File:
  • Angle (deg): Angle from center field (-45° to +45°)
  • Distance (ft): Distance to outfield wall in feet
• Distance Visualization Rays:
  • Green: Success: wall detection with distance measurement
  • Yellow: Warning: field edge detected, but wall not found
  • Red: Failure: no boundary detected

Refinement / Troubleshooting

If measurements seem suboptimal:

• Verify ballpark field clearly visible in image
• Increase / decrease --smooth-level to reweigh data vs smoothness
• Modify Google Maps --zoom and --size for better resolution
• Decrease --min-distance-feet for small parks

Limitations

• Requires clear satellite imagery with visible field boundaries
• Works best on professional/well-defined baseball fields
• Accuracy depends on image resolution and zoom level
• May struggle with obscured boundaries or unusual field layouts