Back to Projects

Archival Digitization Project

Project Hines - WWII Flight Logbook Digitization

A personal archival and data-structuring project focused on converting handwritten WWII-era flight logbooks into verified, searchable spreadsheet records.

Overview

Project Hines is a long-term digitization project built around my great-grandfather’s WWII flight logbooks. The goal is to preserve the original historical records while turning handwritten logbook entries into structured digital data that can be searched, filtered, reviewed, and expanded over time.

The project combines manual transcription, spreadsheet design, image preprocessing, OCR experimentation, and human verification. Because the handwriting, abbreviations, page layout, and historical context are inconsistent, the workflow is designed to prioritize accuracy over full automation.

Current Status

  • Created an initial verified transcription batch ending in October 1939.
  • Structured the first batch into a spreadsheet format for review, correction, and future expansion.
  • Established a baseline workflow for comparing scanned logbook pages against manually verified entries.
  • Continuing to refine the process before scaling the workflow across additional logbook pages.

Project Goal

The end goal is a searchable archive that preserves each flight entry in a consistent format while keeping the original meaning and context of the handwritten records intact. The project is intentionally iterative: each corrected batch improves the structure, validation rules, and future OCR workflow.

Workflow

  • Scan organization: Maintain source images and transcription files in an organized project structure.
  • Image review: Inspect scanned pages for readability, layout issues, handwriting variation, and unclear entries.
  • Manual transcription: Enter flight log data into a structured spreadsheet while preserving dates, locations, aircraft details, flight times, names, and notes where available.
  • Verification: Review unclear entries against the original scan and flag uncertain words or abbreviations for later correction.
  • OCR experimentation: Test preprocessing and OCR tools to assist with extraction, while keeping human review as the accuracy control.
  • Spreadsheet export: Store the cleaned data in a format that can be searched, filtered, corrected, and expanded into a larger archive.

Tools and Technologies

  • Python
  • pandas
  • openpyxl
  • Excel
  • OpenCV
  • Tesseract
  • OCR
  • Git
  • Windows

Technical Focus Areas

  • Data cleaning and spreadsheet structuring
  • Image preprocessing for historical documents
  • OCR testing for difficult handwriting
  • Human-in-the-loop verification
  • Repeatable project organization and documentation

What This Project Demonstrates

Project Hines demonstrates practical problem solving with imperfect real-world data. Unlike a classroom dataset, the logbooks contain inconsistent handwriting, historical abbreviations, variable page quality, and entries that require judgment during transcription. The project shows my ability to break a messy problem into a repeatable technical workflow, document decisions, validate output, and improve the process over time.

It also reflects skills that transfer directly into IT and data-focused roles: attention to detail, structured documentation, troubleshooting, data validation, file organization, and the ability to combine automation with manual quality control.

Next Steps

  • Continue transcribing and verifying additional logbook pages.
  • Refine the spreadsheet structure as new entry types and edge cases appear.
  • Improve image preprocessing to support more reliable OCR testing.
  • Evaluate OCR fallback options for difficult handwriting while keeping manual review in the loop.
  • Build a more navigable archive view once enough verified data has been collected.