Robust Text Line Detection in Historical Documents: Learning and Evaluation Methods

TL;DR

This paper investigates the development of robust, generalizable text line segmentation models for historical documents, emphasizing annotation unification and comprehensive evaluation strategies to improve accuracy across diverse degraded texts.

Contribution

It introduces a unified annotation approach and evaluates three state-of-the-art models, demonstrating improved generalization and segmentation performance on varied historical document datasets.

Findings

Unified annotations enhance segmentation accuracy.

Models trained on diverse datasets generalize well to unseen pages.

Comprehensive evaluation metrics provide better assessment of segmentation quality.

Abstract

Text line segmentation is one of the key steps in historical document understanding. It is challenging due to the variety of fonts, contents, writing styles and the quality of documents that have degraded through the years. In this paper, we address the limitations that currently prevent people from building line segmentation models with a high generalization capacity. We present a study conducted using three state-of-the-art systems Doc-UFCN, dhSegment and ARU-Net and show that it is possible to build generic models trained on a wide variety of historical document datasets that can correctly segment diverse unseen pages. This paper also highlights the importance of the annotations used during training: each existing dataset is annotated differently. We present a unification of the annotations and show its positive impact on the final text recognition results. In this end, we present a complete evaluation strategy using standard pixel-level metrics, object-level ones and introducing goal-oriented metrics.