IDEAS home Printed from https://ideas.repec.org/h/spr/prochp/978-3-032-14099-9_58.html

Generative AI in OCR: A Case Study for License Plate Recognition

In: Recent Advances in Information Systems

Author

Listed:
  • Annita Ngatchou Weiss

    (Aalen University)

  • Lars Guggolz

    (Aalen University)

  • Tim Dahmen

    (Aalen University
    German Research Centre for Artificial Intelligence (DFKI))

  • Christian Koot

    (Aalen University)

Abstract

Accurate character recognition is critical for automatic license plate recognition systems. However, existing optical character recognition (OCR) models often suffer from substitution, addition, and deletion errors, especially under visual artifacts and ambiguous character forms. Single-modal OCR models have limited contextual understanding. In this study, we evaluate whether recent multimodal large language models (LLMs) with vision capabilities, such as GPT-4o and Gemini can reduce character-level recognition errors by leveraging semantic and spatial context. We compare these models against single-modal OCR systems, PaddleOCR, TrOCR, EasyOCR, and Tesseract. Our results show that multimodal LLMs outperformed the single-modal OCR models in terms of overall accuracy (GPT-4o: $$97.6 \pm 0.7$$ 97.6 ± 0.7 %, Gemini: $$83.1 \pm 1.6$$ 83.1 ± 1.6 %). They also exhibit reduced character error rates (CERs), notably in substitution, deletion, and insertion errors. We further show that introducing ensemble inputs to a fine-tuned T5, a generative language model with natural language processing capabilities, as a post-processing step in an OCR pipeline, enables single-modal models to almost match the performance level of Gemini. These findings suggest that the integration of generative models into OCR pipelines can enhance the robustness of license plate recognition systems, particularly under visually challenging conditions.

Suggested Citation

  • Annita Ngatchou Weiss & Lars Guggolz & Tim Dahmen & Christian Koot, 2026. "Generative AI in OCR: A Case Study for License Plate Recognition," Progress in IS, in: Jorge Marx Gómez & Roopesh Kevin Sungkur & Sameerchand Pudaruth & Jan-Hendrik Witte & Gerrit Schuman (ed.), Recent Advances in Information Systems, pages 875-890, Springer.
  • Handle: RePEc:spr:prochp:978-3-032-14099-9_58
    DOI: 10.1007/978-3-032-14099-9_58
    as

    Download full text from publisher

    To our knowledge, this item is not available for download. To find whether it is available, there are three options:
    1. Check below whether another version of this item is available online.
    2. Check on the provider's web page whether it is in fact available.
    3. Perform a
    for a similarly titled item that would be available.

    More about this item

    Keywords

    ;
    ;
    ;
    ;
    ;

    Statistics

    Access and download statistics

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:spr:prochp:978-3-032-14099-9_58. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    We have no bibliographic references for this item. You can help adding them by using this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: Sonal Shukla or Springer Nature Abstracting and Indexing (email available below). General contact details of provider: http://www.springer.com .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.