Advancements and Challenges in Handwritten Text Recognition: A Comprehensive Survey

International audience ; Handwritten Text Recognition (HTR) is essential for digitizinghistorical documents in different kinds of archives. In this study,we introduce a hybrid form archive written in French: the Belfortcivil registers of births. The digitization of these historicaldocuments is challenging due to their unique characteristics suchas writing style variations, overlapped characters and words, andmarginal annotations. The objective of this survey paper is tosummarize research on handwritten text documents and provideresearch directions toward effectively transcribing this Frenchdataset. To achieve this goal, we presented a brief survey ofseveral modern and historical HTR offline systems of differentinternational languages, and the top state-of-the-art contributionsreported of the French language specifically. The survey classifiesthe HTR systems based on techniques employed, datasets used,publication years, and the level of recognition. Furthermore, ananalysis of the systems’ accuracies is presented, highlighting thebest-performing approach. We have also showcased the performance ofsome HTR commercial systems. In addition, this paper presents asummarization of the HTR datasets that publicly available,especially those identified as benchmark datasets in theInternational Conference on Document Analysis and Recognition(ICDAR) and the International Conference on Frontiers inHandwriting Recognition (ICFHR) competitions. This paper,therefore, presents updated state-of-the-art research in HTR andhighlights new directions in the research field.