← Projects
Golfzon Scorecard OCR

Golfzon Scorecard OCR

Automated golf scorecard processing using OCR. Digitize physical scorecards into structured data for tracking and analysis.

PythonOCRReact

Automated golf scorecard processing using OCR. Digitize physical scorecards into structured data for tracking and analysis.

PythonOCRReact
GitHub ↗

Golfzon Scorecard OCR

A tool for automatically extracting and digitizing golf scorecard data from Golfzon simulator screenshots. Upload a photo, get structured data back instantly.

The Problem

Golfzon simulators display scorecards on screen after each round but provide no way to export that data. Running a casual league meant someone manually typing scores from phone screenshots into a spreadsheet after every session. It was slow, error-prone, and nobody wanted to do it.

The Solution

An OCR pipeline that processes Golfzon scorecard images and extracts player names, gross scores, and handicaps into structured data. Built as part of a larger league management system.

How It Works

  1. Image preprocessing. Convert to grayscale, apply CLAHE (contrast-limited adaptive histogram equalization) to normalize screen glare and uneven lighting
  2. OCR extraction. Tesseract for local processing, Google Cloud Vision as a fallback for difficult images
  3. Regex parsing. Extract the consistent PlayerName Score(+/-X) Handicap pattern from raw OCR output
  4. Validation. Score range checks, duplicate detection, manual review UI for edge cases

Accuracy is around 90% on clean photos. The review step before saving covers the remainder without much friction.

Results

Score entry went from a 15-minute spreadsheet process to a 30-second photo upload. The league has been running on this system for several months without issues.