ldas.jp
JA
All tags

TAG

#ocr

68articles

📄The Invisible Text Layer of a Searchable PDF Is Misaligned in Vertical Text — Identifying the Cause (the Source PDF's /Rotate) and Fixing It Without Re-OCR

For a searchable PDF that overlays an OCR invisible text layer on a scanned image, this article identifies the cause of the invisible text drifting on a vertically written document (a double rotation of the coordinate system caused by the source PDF's /Rotate, plus an offset in the line origin) and summarizes how to fix only the invisible text layer without re-running OCR.

pdfocrpythonpymupdf

📄Fixing Unprintable Scanned PDFs Without Breaking the Invisible Text

For PDFs that overlay invisible OCR text on scanned images, a procedure and verification method for unifying page sizes and making JBIG2 images printable (converting to CCITT G4) without breaking the invisible text — using pikepdf and pypdf instead of Ghostscript. Includes the polarity (black/white inversion) pitfall.

pdfpythonocrpikepdf

📚DH Weekly Topics — 2026年6月28日週

DH Weekly Topics — 2026年6月28日週

dh-weeklydigitalhumanitiesデジタルアーカイブocr

📚DH Weekly Topics — 2026年6月21日週

DH Weekly Topics — 2026年6月21日週

dh-weeklydigitalhumanitiesデジタルアーカイブocr

📚DH Weekly Topics — 2026年6月14日週

DH Weekly Topics — 2026年6月14日週

dh-weeklydigitalhumanitiesデジタルアーカイブocr

📚DH Weekly Topics — 2026年6月7日週

DH Weekly Topics — 2026年6月7日週

dh-weeklydigitalhumanitiesデジタルアーカイブocr

📚DH Weekly Topics — 2026年5月31日週

DH Weekly Topics — 2026年5月31日週

dh-weeklydigitalhumanitiesデジタルアーカイブocr

📚DH Weekly Topics — 2026年5月24日週

DH Weekly Topics — 2026年5月24日週

dh-weeklydigitalhumanitiesデジタルアーカイブocr

📜Porting Kuzushiji Character Detection and Classical-Japanese OCR to JS/Next.js Without Python — ONNX Quantization and Library Extraction

A record of porting kuzushiji/historical-document character detection (YOLOv11) and NDL's koten-OCR-Lite (RTMDet + PARSeq) to JS/Next.js with onnxruntime-node — no Python. Covers ONNX export, SAHI, line detection, an int8 quantization accuracy comparison, why it is hard to run on Vercel, and extracting the core into a shared library.

kuzushijiocronnxonnxruntime

📚DH Weekly Topics — 2026年5月17日週

DH Weekly Topics — 2026年5月17日週

dh-weeklydigitalhumanitiesデジタルアーカイブocr

📚DH Weekly Topics — 2026年5月10日週

DH Weekly Topics — 2026年5月10日週

dh-weeklydigitalhumanitiesデジタルアーカイブocr

📜Building a macOS App That Turns a Folder of Page Images Into a Single TEI/XML — From Apple Vision OCR to Developer ID Distribution

I built TEI Scanner, a macOS desktop app that runs Apple's Vision OCR over a folder of scanned page images and emits a single TEI/XML document. This post records the SwiftUI implementation, project management via xcodegen, notarization driven by an App Store Connect API key, and .dmg distribution through GitHub Releases.

macosswiftswiftuivision

Mirador 4.0.0 hides supplementing annotations from the Annotations panel — a `filteredMotivations` gotcha

I delivered IIIF Presentation 3 OCR text annotations with `motivation: "supplementing"`, and they showed up in Annona and other viewers but not in Mirador 4.0.0's Annotations side panel. Reverse-engineering the deployed Mirador bundle revealed that the released default for `config.annotations.filteredMotivations` is `['oa:commenting', 'oa:tagging', 'sc:painting', 'commenting', 'tagging']` — `supplementing` isn't in the allowlist. This post walks through how I found that, the `['commenting', 'supplementing']` array workaround, and the relevant spec / Cookbook references.

iiifmiradorannotationpresentation-api

Comparing NDL Koten OCR-Lite and Cloud Vision API on a Jiaxing Tripitaka 'Mahāprajñāpāramitā Sūtra' — Observations across 105 Images

We applied two OCR engines — Japan's National Diet Library NDL Koten OCR-Lite and Cloud Vision API DOCUMENT_TEXT_DETECTION — to 105 IIIF images of fascicles 571–575 of the Mahāprajñāpāramitā Sūtra in the Jiaxing Tripitaka held by Yūrenja (formerly the Hōonzō of Zōjōji), and compared the patterns of error in their outputs. NDL produced phantom kana lines on 12 pages; Vision picked up color charts, rulers, and shelf labels as if they were body text on all 105.

ocrndl-koten-ocrgoogle-vision-apiiiif

📚DH Weekly Topics — 2026年4月19日週

DH Weekly Topics — 2026年4月19日週

dh-weeklydigitalhumanitiesデジタルアーカイブocr

📚DH Weekly Topics — 2026年4月12日週

DH Weekly Topics — 2026年4月12日週

dh-weeklydigitalhumanitiesデジタルアーカイブocr

📚DH Weekly Topics — 2026年4月5日週

DH Weekly Topics — 2026年4月5日週

dh-weeklydigitalhumanitiesデジタルアーカイブocr

📖Calling NDLOCR-Lite as a Python Library Instead of CLI

How to import and call NDLOCR-Lite, a Japanese OCR engine published by the National Diet Library, directly from a Python script instead of using the CLI.

ndlocrocrpythonjapanese-ocr

KotenOCR Android Development Log — Migrating from Flutter to Kotlin

Development log of building the KotenOCR Android app with Flutter, then migrating to Kotlin Native. Performance benchmarks, API fix for 18x speedup, and framework selection with AI-assisted development.

flutterkotlinonnxruntimeocr

📚DH Weekly Topics — 2026年3月28日週

DH Weekly Topics — 2026年3月28日週

dh-weeklydigitalhumanitiesデジタルアーカイブocr

🔍KotenOCR: Fixing Duplicate Detections in Modern-Script OCR Mode (Adding NMS and Class Filtering)

Fixing duplicate detection results in KotenOCR's modern-script OCR mode (NDLOCR-Lite) by adding NMS and filtering to line_* classes, achieving detection accuracy on par with the reference ndlocr-lite implementation.

iosocrndlonnx

📜KotenOCR v1.3.0: Dual OCR Modes for Classical and Modern Japanese Text

KotenOCR v1.3.0 adds NDLOCR-Lite support for modern printed text OCR alongside the existing classical text (kuzushiji) mode, with one-tap mode switching and parallelized recognition for up to 6.7x speedup.

iosocrkuzushijindl

📚DH Weekly Topics — 2026年3月21日週

DH Weekly Topics — 2026年3月21日週

dh-weeklydigitalhumanitiesデジタルアーカイブocr

⚡Parallelizing OCR Recognition on iOS with Swift Concurrency for up to 6.7x Speedup

Using withThrowingTaskGroup to parallelize OCR recognition in an iOS app, achieving 1.4x to 6.7x speedup depending on the number of detected text regions.

iosswiftconcurrencyocr

📜KotenOCR: An Offline iOS App for Recognizing Classical Japanese Cursive Script

KotenOCR is a free iOS app that runs the NDL Koten OCR-Lite model entirely on-device, enabling offline recognition of kuzushiji (classical Japanese cursive script) from photos.

iosocrkuzushijiswift

✍️Transkribus: AI-Powered Handwritten Text Recognition for Historical Documents

An introduction to Transkribus, an AI-based handwritten text recognition platform supporting 100+ languages, and its applications in Digital Humanities research for transcribing historical documents.

transkribusocrhtrhandwriting

📚DH Weekly Topics — 2026年2月22日週

DH Weekly Topics — 2026年2月22日週

dh-weeklydigitalhumanitiesデジタルアーカイブocr

🐷BDRC Tibetan OCR: Introduction and Implementation Examples of a Tibetan OCR Tool

BDRC Tibetan OCR: Introduction and Implementation Examples of a Tibetan OCR Tool

ocr

📝Azure OpenAI GPT-4 vs Document Intelligence: Comparative Evaluation of Japanese Vertical Text OCR

Azure OpenAI GPT-4 vs Document Intelligence: Comparative Evaluation of Japanese Vertical Text OCR

azureocrllm

🐈LLM-Based Manuscript Paper OCR Performance Comparison: Verification of Vertical Japanese Recognition Accuracy

LLM-Based Manuscript Paper OCR Performance Comparison: Verification of Vertical Japanese Recognition Accuracy

ocrllm

😊Challenges and Solutions for Preserving Order in PDF Transparent Text Extraction

Challenges and solutions for preserving order in PDF transparent text extraction

pdfocr

😸TEI ODD File Customization: A Case Study with NDL Classical Book OCR

TEI ODD File Customization: A Case Study with NDL Classical Book OCR

ocrrngteiroma

👌Development of the NDL Kotenseki OCR-Lite Next.js Version

Development of the NDL Kotenseki OCR-lite Next.js version

nextjsrngocrroma

🌟A Scalable OCR Processing System Using NDL Classical Japanese OCR Lite on Azure Container Apps

A Scalable OCR Processing System Using NDL Classical Japanese OCR Lite on Azure Container Apps

dockerazureocr

🐈Trying DToC: Dynamic Table of Contexts

Trying DToC: Dynamic Table of Contexts

ocrxmlteigemini

🙄Creating TEI/XML Files from IIIF Manifest Files Using NDL Kotenseki OCR-Lite

Creating TEI/XML files from IIIF manifest files using NDL Kotenseki OCR-Lite

ocrhuggingfaceiiiftei

😸Part 2: Creating Annotated IIIF Manifest Files and TEI/XML Files Using NDL Classical Book OCR-Lite

Part 2: Creating Annotated IIIF Manifest Files and TEI/XML Files Using NDL Classical Book OCR-Lite

nextjsocrmonorepoiiif

🎉A Program to Create TEI/XML Files with OCR Results from IIIF Manifest Files

A program to create TEI/XML files with OCR results from IIIF manifest files

colabocriiiftei

📑Created a Similar Text Search App for the Koui Genji Monogatari

Created a Similar Text Search App for the Koui Genji Monogatari

pythonocrhuggingfacetei

🦔Building an NDLOCR Gradio App Using Azure Virtual Machines

Building an NDLOCR Gradio App Using Azure Virtual Machines

azuregpuocrndl

🎉Created a Gradio App to Try ndlocr_cli (NDLOCR ver.2.1) Application

Created a Gradio app to try ndlocr_cli (NDLOCR ver.2.1) application

azureocrndlgradio

🔥Building a Gradio App Using NDL Kotenseki OCR-Lite

Building a Gradio app using NDL Kotenseki OCR-Lite

dockerocrgradio

🔥Using NDL Classical Book OCR-Lite (ndlkotenocr-lite) on Mac OS

Using NDL Classical Book OCR-Lite (ndlkotenocr-lite) on Mac OS

ocrndl

🙆Creating a Transparent Text PDF from a Single Page Using Google Cloud Vision API

Creating a transparent text PDF from a single page using Google Cloud Vision API

pythonpdfocr

📖Mirador Repository with Vertical Text Support for the Text Overlay Plugin

Mirador Repository with Vertical Text Support for the Text Overlay Plugin

ocriiifmirador

😺Applying Google Cloud Vision to Image Files to Create IIIF Manifests and TEI/XML Files

Applying Google Cloud Vision to Image Files to Create IIIF Manifests and TEI/XML Files

pythonocriiiftei

📘Handling Shared Memory Shortage When Running ndlocr_cli and Other Issues

Handling Shared Memory Shortage When Running ndlocr_cli and Other Issues

ocrndlocr

🔖Disk Space After Installing ndlocr_cli with Docker

Disk space after installing ndlocr_cli with Docker

ocrmdx

📝Created Notebooks Using NDLOCR and NDL Classical Japanese OCR ver.2

Created Notebooks Using NDLOCR and NDL Classical Japanese OCR ver.2

colabocriiifndl

💭Running NDL Classical Japanese OCR on mdx

Running NDL Classical Japanese OCR on mdx

ocrndlmdxjp

📝Mirador 3 Plugin Development: Adding Vertical Text Support to the Text Overlay Plugin

Mirador 3 plugin development: adding vertical text support to the Text Overlay plugin

ocrxmliiifmirador

📌About ALTO (Analyzed Layout and Text Object) XML

About ALTO (Analyzed Layout and Text Object) XML

ocrxmlalto

💭Bug Fixes and Feature Additions to the NDL Classical Book OCR Tutorial Using Google Colab

Bug fixes and feature additions to the NDL classical book OCR tutorial using Google Colab

pythoncolabocriiif

🐡Web Application for NDL Classical Book OCR Using Hugging Face Space

Web Application for NDL Classical Book OCR Using Hugging Face Space

awsocrhuggingface

💡Running NDL Classical Japanese OCR on Amazon EC2 CPU Environment

Running NDL Classical Japanese OCR on Amazon EC2 CPU Environment

awsec2ocr

🙆‍♀️Running NDL Classical Text OCR Using Amazon SageMaker Studio

Running NDL Classical Text OCR Using Amazon SageMaker Studio

awsocr

👋NDL Classical Text OCR Using Google Colab

NDL classical text OCR using Google Colab

pythonmachinelearningocrndlocr

💭Building a Layout Extraction Model Using the NDL-DocL Dataset and YOLOv5

Building a layout extraction model using the NDL-DocL dataset and YOLOv5

colabyoloocrdetection

🌟NDL OCR Now Supports Ruby (Furigana) Text Extraction

NDL OCR Now Supports Ruby (Furigana) Text Extraction

colabocrndl

🐡Created a Video on How to Use the NDLOCR App with Google Colab

Created a Video on How to Use the NDLOCR App with Google Colab

ocrcolab

🐥Running gcv2hocr on Google Colab: Creating Searchable PDFs with Transparent Text Using Google Vision API

Running gcv2hocr on Google Colab: Creating searchable PDFs with transparent text using Google Vision API

colabocrgcv

🐥Created Version 2 of the NDLOCR App Using Google Colab

Created Version 2 of the NDLOCR App Using Google Colab

colabocr

📚Execution Time for NDLOCR Using Google Colab

Execution time for NDLOCR using Google Colab

pythonmachinelearningcolabocr

👌Running NDLOCR App with Google Colab (Image Input and Result Saving via Google Drive)

Running NDLOCR App with Google Colab (Image Input and Result Saving via Google Drive)

machinelearningcolabocr

👋Running the NDLOCR Application Using Google Cloud Platform Compute Engine

Running the NDLOCR Application Using Google Cloud Platform Compute Engine

dockermachinelearningocr

📌Added TEI/XML Download Functionality to the "NDL OCR x IIIF" App

Added TEI/XML Download Functionality to the "NDL OCR x IIIF" App

ocriiiftei

💬[Development Guide] I Created an App to View OCR Results Published by the National Diet Library's Next-Generation Digital Library in an IIIF Viewer

[Development guide] I created an app to view OCR results published by the National Diet Library's Next-Generation Digital Library in an IIIF viewer

awsocriiif

🐥An App for Viewing OCR Results from the NDL "Next-Generation Digital Library" in an IIIF Viewer

An App for Viewing OCR Results from the NDL "Next-Generation Digital Library" in an IIIF Viewer

ocriiif