Comparative Performance of YOLOv8, YOLOv9, YOLOv10, and YOLOv11 for Layout Analysis of Historical Documents Images
Source abstract
Abstract
The digitalization of historical documents is of interest for many reasons, including historical preservation, accessibility, and searchability. One of the main challenges with the digitization of old newspapers involves complex layout analysis, where the content types of the document must be determined. In this context, this paper presents an evaluation of the most recent YOLO methods for the analysis of historical document layouts. Initially, a new dataset called BHN was created and made available, standing out as the first dataset of historical Brazilian newspapers for layout detection. The experiments were held using the YOLOv8, YOLOv9, YOLOv10, and YOLOv11 architectures. For training, validation, and testing of the models, the following historical newspaper datasets were combined: BHN, GBN, and Printed BlaLet GT. Recall, precision, and mean average precision (mAP) were used to evaluate the performance of the models. The results indicate that the best performer was YOLOv8, with a Recalltest of 81% and an mAPtest of 89%. This paper provides insights on the advantages of these models in historical document layout detection and also promotes improvement of document image conversion into editable and accessible formats.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Advanced Neural Network Applications · Industrial Vision Systems and Defect Detection · 3D Surveying and Cultural Heritage
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
PAVIC Laboratory, University of Acre (UFAC), Rio Branco 69915-900, Brazil
Funding
PAVIC Laboratory, University of Acre, Brazil, benefited from SUFRAMA fiscal incentives under Brazilian Law
References (44)
Show source-provided references
1. A brief review of state-of-the-art object detectors on benchmark document images datasets10.1007/s10032-023-00431-0 ↗
2. Library digitization projects, issues and guidelines: A survey of the literature10.1108/07378830610669637 ↗
3. Barlindhaug, G. (2022). Artificial Intelligence and the Preservation of Historic Documents, University of Akron Press.10.35492/docam/9/2/9 ↗
4. Biffi, S. (2016). Document Layout Analysis: Segmentation and Classification with Computer Vision and Deep Learning Techniques. [Master’s Thesis, ING-Scuola di Ingegneria Industriale e dell’Informazione].
5. Optical character recognition-A review
6. Araújo, A.B. (2019). Análise de layout de página em jornais históricos germano-brasileiros. [Master’s Thesis, UFPR-Federal University of Paraná].
7. Jannidis, F., Kohle, H., and Rehbein, M. (2017). Digital Humanities, Springer.10.1007/978-3-476-05446-3 ↗
8. Zhu, W., Sokhandan, N., Yang, G., Martin, S., and Sathyanarayana, S. (March, January 22). DocBed: A multi-stage OCR solution for documents with complex layouts. Proceedings of the AAAI Conference on Artificial Intelligence, Online.
9. An adaptive over-split and merge algorithm for page segmentation10.1016/j.patrec.2016.06.011 ↗
10. Shehzadi, T., Stricker, D., and Afzal, M.Z. (September, January 30). A hybrid approach for document layout analysis in document images. Proceedings of the International Conference on Document Analysis and Recognition, Athens, Greece.
11. A review of YOLO object detection algorithms based on deep learning10.54097/fcis.v4i2.9730 ↗
12. Lavanya, G., and Pande, S.D. (2024). Enhancing Real-time Object Detection with YOLO Algorithm. EAI Endorsed Trans. Internet Things, 10.10.4108/eetiot.4541 ↗
13. Wang, C.Y., and Liao, H.Y.M. (2024). YOLOv1 to YOLOv10: The fastest and most accurate real-time object detection systems. APSIPA Trans. Signal Inf. Process., 13.10.1561/116.20240058 ↗
14. Saha, R., Mondal, A., and Jawahar, C. (2019, January 20–25). Graphical object detection in document images. Proceedings of the 2019 International Conference on Document Analysis and Recognition (ICDAR), Sydney, Australia.10.1109/ICDAR.2019.00018 ↗
15. Alaasam, R., Kurar, B., and El-Sana, J. (2019, January 20–25). Layout analysis on challenging historical arabic manuscripts using siamese network. Proceedings of the 2019 International Conference on Document Analysis and Recognition (ICDAR), Sydney, Australia.10.1109/ICDAR.2019.00123 ↗
16. Xu, Y., Li, M., Cui, L., Huang, S., Wei, F., and Zhou, M. (2020, January 6–10). Layoutlm: Pre-training of text and layout for document image understanding. Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, Virtual.10.1145/3394486.3403172 ↗
17. Appalaraju, S., Jasani, B., Kota, B.U., Xie, Y., and Manmatha, R. (2021, January 10–17). Docformer: End-to-end transformer for document understanding. Proceedings of the IEEE/CVF International Conference on Computer Vision, Montreal, QC, Canada.10.1109/ICCV48922.2021.00103 ↗
18. Huang, Y., Lv, T., Cui, L., Lu, Y., and Wei, F. (2022, January 10–14). Layoutlmv3: Pre-training for document ai with unified text and image masking. Proceedings of the 30th ACM International Conference on Multimedia, Lisbon, Portugal.10.1145/3503161.3548112 ↗
19. Chen, S., Sun, P., Song, Y., and Luo, P. (2023, January 2–3). Diffusiondet: Diffusion model for object detection. Proceedings of the IEEE/CVF International Conference on Computer Vision, Paris, France.10.1109/ICCV51070.2023.01816 ↗
20. Zhao, Y., Lv, W., Xu, S., Wei, J., Wang, G., Dang, Q., Liu, Y., and Chen, J. (2024, January 17–18). Detrs beat yolos on real-time object detection. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA.10.1109/CVPR52733.2024.01605 ↗
21. Zhao, Z., Kang, H., Wang, B., and He, C. (2024). Doclayout-yolo: Enhancing document layout analysis through diverse synthetic data and global-to-local adaptive perception. arXiv.
22. Chai, M., Shen, Z., Zhang, C., Zhang, Y., Wang, X., Dou, S., Kang, J., Zhang, J., and Zhang, Q. (2024). DocFusion: A Unified Framework for Document Parsing Tasks. arXiv.
23. A COMPREHENSIVE EVALUATION OF YOLOv5s AND YOLOv5m FOR DOCUMENT LAYOUT ANALYSIS
24. The YOLO model that still excels in document layout analysis10.1007/s11760-023-02838-y ↗
25. YOLOLayout: Multi-Scale Cross Fusion Former for Document Layout Analysis
26. Fleischhacker, D., Goederle, W., and Kern, R. (2024). Improving OCR Quality in 19th Century Historical Documents Using a Combined Machine Learning Based Approach. arXiv.
27. Chen, H.C., Wu, L., and Zhang, Y. (2025). EDocNet: Efficient Datasheet Layout Analysis Based on Focus and Global Knowledge Distillation. arXiv.
28. Ströbel, P., and Clematide, S. (2019, January 9–12). Improving OCR of Black Letter in Historical Newspapers: The Unreasonable Effectiveness of HTR Models on Low-Resolution Images. Proceedings of the Digital Humanities 2019 (DH2019), Utrecht, The Netherlands.
29. Chollet, F. (2021). Deep Learning with Python, Simon and Schuster.
30. Achieving optimal process design for minimizing porosity in additive manufacturing of Inconel 718 using a deep learning-based pore detection approach10.1007/s00170-022-09372-0 ↗
31. Sulzbach, L. (2022). Building a Dataset of Late Modern German-Brazilian Newspaper Advertisement Pages for Layout and Font Recognition. [Bachelor’s Thesis, UFPR-Federal University of Paraná].
32. A comprehensive review of yolo architectures in computer vision: From yolov1 to yolov8 and yolo-nas10.3390/make5040083 ↗
33. Wang, C.Y., Yeh, I.H., and Mark Liao, H.Y. (October, January 29). Yolov9: Learning what you want to learn using programmable gradient information. Proceedings of the European Conference on Computer Vision, Milan, Italy.
34. Wang, A., Chen, H., Liu, L., Chen, K., Lin, Z., Han, J., and Ding, G. (2024). Yolov10: Real-time end-to-end object detection. arXiv.
35. Sapkota, R., Meng, Z., Churuvija, M., Du, X., Ma, Z., and Karkee, M. (2024). Comprehensive performance evaluation of yolo11, yolov10, yolov9 and yolov8 on detecting and counting fruitlet in complex orchard environments. arXiv.10.32388/E9Y7XI ↗
36. Khanam, R., and Hussain, M. (2024). YOLOv11: An Overview of the Key Architectural Enhancements. arXiv.
37. Hyperparameter optimization of YOLOv8 for smoke and wildfire detection: Implications for agricultural and environmental safety
38. Ji, X., Wan, J., Zhang, Y., Yang, T., Yang, Q., and Liao, J. (2024). Hydrospot-Yolo: Real-Time Water Hyacinth Detection in Complex Water Environments, Elsevier.10.2139/ssrn.5052244 ↗
39. Neupane, C., Walsh, K.B., Goulart, R., and Koirala, A. (2024). Developing Machine Vision in Tree-Fruit Applications—Fruit Count, Fruit Size and Branch Avoidance in Automated Harvesting. Sensors, 24.10.3390/s24175593 ↗
40. YOLOv8-FDD: A real-time vehicle detection method based on improved YOLOv810.1109/ACCESS.2024.3453298 ↗
41. Lee, Y.S., Prakash Patil, M., Kim, J.G., Seo, Y.B., Ahn, D.H., and Kim, G.D. (2025). Hyperparameter Optimization for Tomato Leaf Disease Recognition Based on YOLOv11m. Plants, 14.10.3390/plants14050653 ↗
42. Padilla, R., Netto, S.L., and Da Silva, E.A. (2020, January 1–3). A survey on performance metrics for object-detection algorithms. Proceedings of the 2020 international conference on systems, signals and image processing (IWSSIP), Niterói, Brazil.10.1109/IWSSIP48289.2020.9145130 ↗
43. Terven, J., Cordova-Esparza, D.M., Ramirez-Pedraza, A., and Chavez-Urbiola, E.A. (2023). Loss functions and metrics in deep learning. A review. arXiv.
44. Wang, B. (2022). A parallel implementation of computing mean average precision. arXiv.