Skip to content

leonbeier/Scratch_Detection

Repository files navigation

This directory contains cleaned visualization scripts for sharing.

What I changed

  • Removed user-specific absolute paths.
  • Scripts now accept CLI arguments and read SCRATCH_MODEL env var.
  • No private Roboflow keys or other secrets are present.

Quick usage

Run with defaults (expects model.onnx and video.mp4 in the current folder):

python visualize_scratches.py

Provide explicit paths:

python visualize_scratches.py --model path/to/model.onnx --video path/to/input.mp4 --output out.mp4 --results results.json

Or set the environment variable for the default model:

Windows PowerShell

$env:SCRATCH_MODEL = "path\to\model.onnx"
python visualize_scratches.py

Unix / macOS

export SCRATCH_MODEL=path/to/model.onnx
python visualize_scratches.py

Notes

  • If you want different defaults for the output filenames, pass --output and --results.
  • The scripts do not contain any API keys or external credentials. If your workflow requires Roboflow or similar, provide keys outside the script (env vars or a secrets manager).

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages