Page Layout Analysis System for Unconstrained Historic Documents

Kodym, Oldřich; Hradiš, Michal

doi:10.1007/978-3-030-86331-9_32

Cited by 14 publications

(4 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(Gupta et al, 2015). All of this makes optical character recognition (OCR) difficult and motivates research into the pre-processing of images to eliminate noise (Neji et al, 2024), or the training of character recognition models specialized in Gothic and round letters using machine learning based on neural networks (Lacasta et al, 2022;Kodym et al, 2021).…”

Section: State Of the Artmentioning

confidence: 99%

Primeros pasos hacia una plataforma para el análisis del patrimonio documental de derecho civil

Neji,

Nogueras-Iso,

García-Marco

et al. 2024

Scire

View full text Add to dashboard Cite

El patrimonio documental sobre derecho civil es un importante activo cuyo estudio permite conocer el contexto político, social y cultural de la época a la que se refieren los documentos históricos. Este trabajo presenta el diseño de un prototipo de plataforma para apoyar a los investigadores en el análisis del patrimonio documental sobre derecho civil. Nuestra plataforma consiste en crear una versión en línea de estos materiales, haciéndolos más accesibles. La plataforma proporciona asistentes automáticos para la transcripción, traducción y extracción de elementos de información específicos asociados a conceptos de derecho civil (voces) como citas a fuentes externas y entidades con nombre (lugares, personas y organizaciones) para identificar mejor su contexto. La viabilidad de esta plataforma se ha puesto a prueba con el tratamiento de una obra doctrinal escrita por Miguel del Molino, un conocido experto en derecho civil del siglo XV en el reino de Aragón.

show abstract

Section: State Of the Artmentioning

confidence: 99%

Primeros pasos hacia una plataforma para el análisis del patrimonio documental de derecho civil

Neji,

Nogueras-Iso,

García-Marco

et al. 2024

Scire

View full text Add to dashboard Cite

show abstract

“…The documents contain 9 languages and 10 font families. We randomly sampled 1.3M text line images from the original collection 5ehJ bGk E p270bH 0njh iFQ0Sk 2zEqWuQ jbqp wE Mvxc5 hzrb8 zm1wogV with ParseNet [18] and we translated their transcriptions into additional nine synthetic TS. Thus, each text line has ten transcription versions where one is the original one.…”

Section: Synthetic Transcription Styles Experimentsmentioning

confidence: 99%

“…We represent a document by a single TSI, as we assume that each document was transcribed by a single organization using a consistent transcription style, but we have no information on how or where the documents were transcribed. We randomly sampled 1.08M text lines from 2445 randomly picked DTA documents with ParseNet [18]. The number of lines per document ranges from 200 to 3029.…”

Section: Real Transcription Styles Experimentsmentioning

confidence: 99%

TS-Net: OCR Trained to Switch Between Text Transcription Styles

Jan¹,

Hradiš²

2021

Preprint

Self Cite

View full text Add to dashboard Cite

Users of OCR systems, from different institutions and scientific disciplines, prefer and produce different transcription styles. This presents a problem for training of consistent text recognition neural networks on real-world data. We propose to extend existing text recognition networks with a Transcription Style Block (TSB) which can learn from data to switch between multiple transcription styles without any explicit knowledge of transcription rules. TSB is an adaptive instance normalization conditioned by identifiers representing consistently transcribed documents (e.g. single document, documents by a single transcriber, or an institution). We show that TSB is able to learn completely different transcription styles in controlled experiments on artificial data, it improves text recognition accuracy on large-scale real-world data, and it learns semantically meaningful transcription style embeddings. We also show how TSB can efficiently adapt to transcription styles of new documents from transcriptions of only a few text lines.

show abstract

“…Several research explorations were carried out on complex documents [2,11], whereas some others focused on more complex historical documents, which may implicate challenging handwritten manuscripts in different languages [12][13][14]. Moreover, DLA can be an even more challenging task when applied to historical handwritten documents with highly unconstrained structure and complex page layouts [15,16], as in ancient/historical Arabic manuscripts [17][18][19].…”

Section: Introductionmentioning

confidence: 99%

Semantic Document Layout Analysis of Handwritten Manuscripts

Jaha¹

2023

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

A document layout can be more informative than merely a document's visual and structural appearance. Thus, document layout analysis (DLA) is considered a necessary prerequisite for advanced processing and detailed document image analysis to be further used in several applications and different objectives. This research extends the traditional approaches of DLA and introduces the concept of semantic document layout analysis (SDLA) by proposing a novel framework for semantic layout analysis and characterization of handwritten manuscripts. The proposed SDLA approach enables the derivation of implicit information and semantic characteristics, which can be effectively utilized in dozens of practical applications for various purposes, in a way bridging the semantic gap and providing more understandable high-level document image analysis and more invariant characterization via absolute and relative labeling. This approach is validated and evaluated on a large dataset of Arabic handwritten manuscripts comprising complex layouts. The experimental work shows promising results in terms of accurate and effective semantic characteristic-based clustering and retrieval of handwritten manuscripts. It also indicates the expected efficacy of using the capabilities of the proposed approach in automating and facilitating many functional, reallife tasks such as effort estimation and pricing of transcription or typing of such complex manuscripts.

show abstract

Page Layout Analysis System for Unconstrained Historic Documents

Cited by 14 publications

References 22 publications

Primeros pasos hacia una plataforma para el análisis del patrimonio documental de derecho civil

Primeros pasos hacia una plataforma para el análisis del patrimonio documental de derecho civil

TS-Net: OCR Trained to Switch Between Text Transcription Styles

Semantic Document Layout Analysis of Handwritten Manuscripts

Contact Info

Product

Resources

About